SSD Price Drops Signaling End of Spinning Media?

gjt writes "When Intel and OCZ recently announced new 'affordable' Solid State Disk drives — offering a meager 32-40GB — we initially yawned. But, then we took a closer look at the press releases and the in-progress research and development in SSD technology and opened our eyes. While the new drives aren't affordable on a cost per gigabyte basis for everyone, it does set a precedent — and most importantly a barometer price of $100. And it really does start the death clock for hard drive technology."

Comments

...Or an arms race

[Dan East]

I think HDD will continue to stay enough ahead of SSD in raw capacity that it will stay relevant for a long time. When SSD is affordable at 200 GB then HDD will already be affordable at 2 TB, etc.

Re:...Or an arms race

[Drethon]

HDD is the new tape drive and SSD is the new HDD?

Re:...Or an arms race

[MindlessAutomata]

200GB ought to be enough for anybody.

Re:...Or an arms race

[ircmaxell]

Well, so then use them for appropriate uses. I don't need 2TB on my laptop (I barely need 40gb). But on my home file server, I could use the spinning disks for brute capacity. So perhaps what we may start seeing is more and more computers shipped with a 20 or 40gb SSD boot disk with a 500gb or 1TB "data disk"... But to say that spinning disks will go away is kinda short sighted...

Re:...Or an arms race

[91degrees]

But seek times suck. Tapes just have their niche. I wonder if disks will.

Re:...Or an arms race

[0123456]

Essentially: I expect my next PC to have an SSD for important program files and data, and HDD(s) for big data files which don't need fast random access (e.g. video files). Or I'll offload them to an OpenSolaris server with a bunch of HDDs in a RAIDZ.

The idea that cheaper SSDs will kill HDD is silly when most peoples' storage needs expand to meet whatever they can afford to buy. Certainly they are likely to kill HDDs in simple home and office systems, but for everything else HDDs will continue to be vastly cheaper for at least the next decade or so.

When the latest game I installed on my PC wanted 20GB of disk space, a $100 200GB SSD won't last long.

Re:...Or an arms race

[LWATCDR]

SSDs will replace all the small hard drives.
When you get down to small enough drive SSDs will be cheaper per Gig than HDs.
Right now you can buy a 1TB drive for right around $90.
But you can not buy a 5ooGB drive for $45 or a 250GB drive for $22.50. There is a limit to how cheap you can make a harddrive.
At some point SSDs in the 120Gb range will be cheaper than spinning platters. It is probably close right now.
When that happens you will see SSDs replace HDs in that range. That range will keep creeping up and up.
So HDDs will be what you get when you need a lot of storage. Maybe they will eventually be used only for externals and NASs.
Eventually 1 TB SSDs will be cheaper than HDDs but for all I know we will have 100TB HDs for $90.
BTW as someone that paid several hundred dollars for a 30MB HD in 1984 the idea of a sub hundred dollar 100TB HDD just seems like a matter of time.

Re:...Or an arms race

[jollyreaper]

I think HDD will continue to stay enough ahead of SSD in raw capacity that it will stay relevant for a long time. When SSD is affordable at 200 GB then HDD will already be affordable at 2 TB, etc.

It will be a long, protracted death. I anticipate seeing HDD for mass storage for another decade bare minimum. SSD will be for the stuff that has to be fast, HDD will be for the stuff that can be slower but SSD is too expensive for. (someone look this up in ten years and laugh at me.)

Re:...Or an arms race

[amorsen]

Tapes don't really have a niche anymore. They're too slow and inconvenient even for backups, and tape robots are not that much cheaper than JBOD.

If only there was a decent free backup system which treated disks as disks instead of pretending they're tape. Bacula and Amanda both suck for to-disk backups.

Re:...Or an arms race

[Sperbels]

...and Floppies are the new punch cards...and punch cards are the new abacuses...and abacuses are the new ...what? Fingers and toes?

Re:...Or an arms race

[exasperation]

I think HDD will continue to stay enough ahead of SSD in raw capacity that it will stay relevant for a long time. When SSD is affordable at 200 GB then HDD will already be affordable at 2 TB, etc.

Ah, but when 200 GB of storage is $20, no hard drive will ever be able to be that cheap. There is a fixed minimum cost for building a hard drive. Spindle, motor, etc. It's about $70. When "enough storage" for the average user, let's say 200 GB costs less than that base cost, almost all new storage sold will be SSD devices due to their overall advantages, especially in a battery-powered machine (which are the majority of all computers sold today).

This will completely gut the market for hard drives and R&D into them will cease. All money will move to SSDs and they will improve even more rapidly.

Re:...Or an arms race

[pitchpipe]

...and Floppies are the new punch cards...and punch cards are the new abacuses...and abacuses are the new ...what? Fingers and toes?

And the middle finger is the new raised eyebrow.

Re:...Or an arms race

[0123456]

This will completely gut the market for hard drives and R&D into them will cease. All money will move to SSDs and they will improve even more rapidly.

Indeed: no-one will ever need more than 200GB of storage.

Re:...Or an arms race

[Hadlock]

I wonder how long it will be before SSDs lose the traditional 3.5" form factor. There's no reason why you couldn't say, drop the guts into a PCI form factor. That cast aluminum enclosure is probably $3-5 of a product that probably costs $45 to make. With less heat and mass requirements it's likely we'll start seeing naked chips on a breadboard to save 8-9% of the manufacturing cost.

Re:...Or an arms race

[MrNemesis]

Oddly enough... increasingly so.

I'm lucky enough to have a spanking new SAN at work, a bunch of iSCSI nodes with dumb discs and smart software. We tier storage over SSD, 10k SAS and 7.2k SATA, the SSD mostly being utilised for performance-critical databases. SATA discs we reserve for file server LUNs, low-I/O VMs (most of them) and the intermediate tier of VM -> LUN -> Tape backups.

As an aside, we choose to use RAID1 for SSD LUNs, as the RAID controllers can't keep up with the parity calculations and we saw a ~40% drop in IOPS when we used RAID6.

Re:...Or an arms race

[Dan Ost]

+1 Insightful

Excluding niche uses, this is exactly how the mainstream transition from platters to SSD will happen.

Re:...Or an arms race

[jandrese]

They still have a niche for people who pull the tapes out and ship them offsite everyday. HDD hot swap connectors aren't really meant to be used every day, and they weigh a lot.

Re:...Or an arms race

[Hatta]

By that logic we should have brand new super cheap CRTs everywhere.

Re:...Or an arms race

[sconeu]

Oh yeah. Right. I really want to access my main storage over a 768Kb/384Kb connection.

No fucking chance until broadband is widely accessible. And of course, your machine is useless without a net connection (e.g. travelling).

Re:...Or an arms race

[MrNemesis]

Not really - whilst spinning platters are awesome for nearline storage, they're not nearly reliable enough to be used as a backup solution IMHO. Stick a bunch of caddied HDD's in a cupboard and leave them for a year and you can practically guarantee one of them will fail to spin up. Tapes will remain in the backup for the foreseeable future because, TTBOMK, there's no other media that's reliable enough over the long term.

Big-assed RAID arrays used for backup storage are all well and good, especially in a small or low-budget environment, but enterprises use tapes for a reason.

Re:...Or an arms race

[Drethon]

And 0100 is the new middle finger? Glad that hasn't seemed to happen...

Re:...Or an arms race

[Anonymous Coward]

A lot of things can change in 10 years. Just think back to the year 2000: computers only ran at 2-3 GHz and Linux was just getting into the mainstream!

Re:...Or an arms race

[SWPadnos]

(most SSD are 2.5", not 3.5")

PCIe "hard drives" already exist.

Here's a 1TB model: http://www.newegg.com/Product/Product.aspx?Item=N82E16820227500
There are others in 250GB, 256GB, and 512GB capacities.

I doubt that the cost goes down much though. The PCIe interface chip isn't free, and neither is the card bracket. The PC board itself is also much larger, and has to be thicker than those used on most hard drives. The cost differences are probably a wash.

Re:...Or an arms race

[Gruturo]

There is a fixed minimum cost for building a hard drive. Spindle, motor, etc. It's about $70.

Not quite. A shop near my house (Rome, Italy) has 320GB drives on sale at 35.4 EUR (roughly $48) including 20% sales tax - and this is just the first I bothered to check, it's a street price, it includes their own profit, and it's a 3.5" unit. When 3.5" go obsolete once and for all, the 2.5" drives will stop costing a premium and actually become cheaper, most likely.

So - while there definitely is a price level where mechanical units stop making sense, it's nowhere near $70 and probably it will keep shrinking over time.
Anyway - the entire point is moot. A sum of other factors (weight, power consumption, heat generation and tolerance, shock tolerance) will most likely push hard disks away in the lower capacity ranges.

Re:...Or an arms race

[MrNemesis]

I reckon we'll need a couple more years of SSD acceptance before this becomes mainstream - for one thing, we'll need an OS-agnostic method of using PCIe cards as bootable block devices, which will probably take a while to work out. Heck, I'm not even sure if FusionIO is bootable yet (I think they're working on it). And then there's all those filesystems in use that all assume they're on spinning discs.

SSD will be made of awesome when this happens though - SATA has been a bottleneck for flash for quite a while now, and removing a) the slow-assed bus and b) the complex disc controller overhead and the futzing around that the OS does will give gargantuan speed improvements. I expect there'll be a shift like there was from MFM to IDE - move the control of the hardware away from the OS and into a smart hardware controller that embedded with the device. SATA is just used now because it's common and universally supported by pretty much everything - crucial at the "early adopter" stage.

Re:...Or an arms race

[amorsen]

True, but 10Gbps fiber is helping a lot in that problem space.

Re:...Or an arms race

[socrplayr813]

The vast majority of SSDs are 2.5" or 1.8", and there are PCIe SSDs out there. I can't speak to the quality and cost, but it's not a new idea:

http://www.newegg.com/Product/ProductList.aspx?Submit=Property&Subcategory=636&Description=&Type=&N=2010150636&srchInDesc=&MinPrice=&MaxPrice=&PropertyCodeValue=4213%3A46884&PropertyCodeValue=4213%3A47725

Re:...Or an arms race

[Weeksauce]

In addition to your perspective, I don't see them taking off until storage capacity is in line with traditional HDD. Will the average consumer recognize the benefits of an SSD or will they say ohh this drives got 10x the capacity of that one, why on earth would I want to pay more for the small one?

Re:...Or an arms race

[LWATCDR]

What you will probably see soon is a new mini-pci standard. The current mini-pci standard has a USB port as part of the standard. Once they start putting USB 3.0 on there it will be trivial to put a USB 3.0 Flash drive chip on a card. You may even see a card that is both an SSD and WiFi which would be great for netbook makers.

Re:...Or an arms race

[Hadlock]

Oh, I was just talking about something that mounted in the PCI slot and was held in with a single screw like an old school sound card. There'd be no pin connectors interfacing directly with the motherboard; it'd still have a SATA jack to wire it to the motherboard.
 
Hell, there's no reason why they couldn't just integrate a 20 or 40gb SSD right into the motherboard. Talk about a microcomputer! Lenovo has some pretty tiny nettops nowadays, I imagine the physical dimensions of the hard drive more or less doubles the thickness of the unit. With a different form factor they could probably reduce the size of the packaging even further.

Re:...Or an arms race

[Dahamma]

I agree, and it's an even wider gap than you suggest!

HDDs are *already* fairly affordable at 2TB - you can get a 2TB HDD for $150-170. A ~200GB (256GB, actually) SSD is closer to $750. For the price of a 2TB HDD, you're only going to get a 32-64GB SSD.

Then again, data sizes have not kept up with storage increases... once you can store more 12MP photos, lossless music, and 1080p high quality video than you need, there really isn't much point for the average consumer to keep upgrading. Clearly it will be a different story for servers, though. Our total HDD storage at work passed the petabyte level a while ago. That storage cost in SSDs would be pushing 7 figures...

Re:...Or an arms race

[LWATCDR]

Of course the scary thing is now that we have 64 bit processors it could be possible to just map the flash right to the address space. I could see a netbook with a flash drive right on the motherboard.

Re:...Or an arms race

[amorsen]

Isn't magnetic leakage onto nearby tape still a problem? Having to put the tape in a drive and let it wind its way through isn't much fun. Not much of a problem in a tape robot, but then the competition to a tape robot is a disk array where they'll be running a good fraction of the time. It's much easier to do regular verification on a disk array; there's enough bandwidth that you can even do it weekly.

Re:...Or an arms race

[b4upoo]

I like hard drives. I'm used to them. To me electronic memory storage seems best suited for very small, portable units.

Re:...Or an arms race

[MrNemesis]

I believe mainframes already do the whole memory page == disc block thing, but I'm not an expert.

I asked the same question on /. a while back regarding more common operating systems and got this response from m.dillon, which seems to indicate it's not really feasible unless the whole software stack is (very) radically altered, or the performance delta between memory and storage becomes alot smaller.

Re:...Or an arms race

[EvanED]

Ah, but when 200 GB of storage is $20, no hard drive will ever be able to be that cheap. There is a fixed minimum cost for building a hard drive. Spindle, motor, etc. It's about $70.

This page would like a word with you. (Hopefully that link works; if not, it's a page of seven drives from WD, Seagate, and Hitachi for 160 GB drives for $50. The cheapest is $37.99.)

Re:...Or an arms race

[cruff]

Tapes don't really have a niche anymore

Actually they do when you need to store petabytes to exabytes of data with very little power and you can tolerate the latencies due to tape mounts. Current enterprise class tape devices are capable of 100 MB/sec easy. The main problem, as with any storage device, is that they have a limited lifetime and if you wish to keep your data beyond that lifetime, you have to keep oozing the data to new media.

Re:...Or an arms race

[Rhys]

But so what? There is going to be a limit on the # of bits you need to store. Text is already simply noise even on a "small" 32 Gbyte SSD -- heck it is only noise even on the tiny 4 gig primary SSD on the Eee901 I have. Audio, particularly on a laptop isn't going to get any bigger than a mp3 -- or worse case flac -- file is now. 5 megs for a mp3 is starting to be noise to a 32 gig SSD.

So how are we going to keep consuming more and more space? I suspect we'll see another computer capability-stable-price-decrease event like happened a little while ago. The idea of getting a decent laptop brand new for $500, say, 10 years ago was crazytalk. The idea you could even get something substandard new for $300 was just a pipedream. But these days with CULV and netbooks those price points are reality for a pretty nice machine. Sure, the netbook isn't much faster than my 6 year old Pentium-M, but it beats the pants off it in battery life (even when the 'M had a new battery) and is dirt-cheap.

Put another way: I'm not convinced I'll really ever want to keep over 2 terabytes of data. In somewhere around 5 years of casually taking digital pictures I've only shot about 6 gigs of photos -- with over half of them not being keeper types (pics for ebay/craigslist, etc). My and my wife's whole ripped music collection is a whopping 29 Gbyte (small enough I could fit Linux and the .oggs on one of said SSDs). Even turing that into .flacs with a re-rip that is likely to sit at what, 150 Gbyte give or take? Add to that another 50 years of photos (at twice the size they are now) and some music collection growth I'm looking at maybe 300-500 Gbyte of storage.

So why do I even need the 2 Tbyte drive? If the SSD hits 1 Tbyte and is affordable (and fast), why would I consider paying similar pricing for space I'll never use on a 100 Tbyte HDD?

Re:...Or an arms race

[jernejk]

Mode this guy up. I'm only buying next laptop when they have dual disk bays, one for SSD and one for HDD. Or, HDD with integrated SSD would also be nice.

Re:...Or an arms race

[MrNemesis]

Dunno about magentic leakage as tapes aren't really my area (mainly because I hate the damned things :))

Personally, I don't really think people using a single tape drive without a robot/library count as "enterprisey" enough for my blood - if I had a single drive, I too would use a nearline HDD-based storage system to keep the last few versions close at hand so there's no need for the rigamarole of spooling through a tape to restore just one file. We implmented just such a "on the cheap" system in a previous job some years ago - we backed up to tape directly from the file server, but also maintained a backup server that rsynced the files over every hour. It used hardlinks to allow us to keep several faux-snapshots of the entire system whilst only actually keeping one instance of every unique file, giving us about 4 months of instant restore-any-file-you-like in just 2x the amount of space in use on the file server. Easy to set up and saved us a fortune on backup infrastructure.

Re:...Or an arms race

[dgatwood]

I think part of the confusion here is that you're confusing the word "backup" with the word "archive". They're not the same thing, though a lot of IT people try to use backups as archives, which is why folks cling to archaic technology like tape.

Hard-drives aren't intended for long-term storage, but that's okay because backups don't have to last long-term. Backups are short-term, by definition. They just have to last long enough to guarantee that you have two or three complete backups of the data, preferably with a couple of those backups off-site and offline at any given time.

Archives, by contrast, provide the ability to fetch an ancient version of some file. Although those can be implemented using tape, they can also be implemented through a version control system. Doing so has the advantage of more rapid availability of the old data, lower overhead (you don't have to call somebody from IT to go digging through a tape vault for an ancient backup tape), and eliminating the need for long-term storage entirely because now even the old versions of the files are being backed up regularly---backups that really only need to last until the drives have their next turn in the backup rotation....

Re:...Or an arms race

[Minwee]

it'd still have a SATA jack to wire it to the motherboard.

Because otherwise it might run too fast? The SATA interface is the big bottleneck holding SSD speeds down. If you put the whole thing directly on the PCIe bus it would be a lot happier. Fitting a fast SSD into a PCIe slot and then tying it to SATA is just cruel.

Re:...Or an arms race

[MrNemesis]

Indeed, yes, you're correct, alot of this is all about the terminology - I tend to conflate the terms myself, but then the distinction between the two can be very murky. True, most of the time a backup of, say, a commonly used database will be next to useless of it's more than a month or two old, but I've run into scenarios where we've had silent corruption in a (thankfully infrequently used) DB that the DBAs didn't notice for nearly six months, which resulted in us pulling the archival tape and restoring from that.

Personally, I wish we had the money for a nearline HDD backup or a VTL in my current job, but seeing as we *have* to have tape for archival reasons and we can turn most restores around fairly quickly, it's not seen as a necessary expense.

Re:...Or an arms race

[Grishnakh]

Not quite. No matter how much data they cram onto a platter, HDs still have the disadvantage of poor access time; their access times haven't changed significantly in a decade or so, even though capacities have gone up 1000-fold. SSDs by their solid-state nature have much faster access times, so they're better for frequently-accessed data.

I think we will start seeing a lot of systems with SSDs for the OS and HDs for user data, media (e.g. video files and MP3s), etc. Even 200GB SSDs won't be enough for all the HD videos and other crap people will have on their systems, but it's certainly more than enough for even the most bloated MS OS and applications.

Re:...Or an arms race

[Jay L]

I thought the same thing, which is why I bought a used 10-tape Exabyte library for my home network. And then I found CrashPlan and its ilk (MozyBackup, etc), and realized:

Statistically, yes, tapes are more reliable than disk - but not perfectly reliable. Which means you need to check them every once in a while to make sure you still have the backups you think you do. And if you're going to fetch the tapes once in a while, it's not any harder to fetch the disks. And if you're fetching the disks every once in a while, it's not any harder (once you've written the software) to maintain the equivalent of a nearline RAID array; any one platter can fail without you losing a byte of data. Now, disks are MORE reliable than your never-verified tapes, and cheaper too.

Oh, and once you're storing your archives on disk, you can do things like automatic x-deltas, versioning, pruning to a grandfather-father-son history, etc. as part of automatic maintenance. And you can maintain eventual consistency with off-site storage. And the list goes on.

Re:...Or an arms race

[Rhys]

You mean like the Eee 701 did?

They'd be better off getting a pseudo-memory standard like Sun is -- reusing the laptop dimm formfactor for a ssd. Or even just using a regular old ddr1 slot with fairly few of the contacts actually live. The size of a dimm is a lot closer to the size of a current-gen 2.5 inch SSD than PCI or PCI-E cards (usually) are.

Re:...Or an arms race

[LWATCDR]

He is correct but that is not what I was thinking.
The OS would know that it is flash and protect it from general access. The applications would still access it through a device driver but instead of going through an IO chip and the sata bus the OS would handle the IO on the memory bus.
The way things happen now is that when you do a disk read the data is copied from the drive across the bus and into the controller chip. The controller chip then does a direct memory access to a block of memory that you have told it to use as a buffer. This is a simple explanation of how it works BTW.
If you just mapped the flash to part of the address space the programs would never know the difference.
What would change is that when you did a read the data would be copied from the slower flash to the ram buffer all across the memory bus.
Now what I don't know is if that would cause issues with having the two different speeds of memory or not. It could be that it is faster to not do all of the IO on the Address bus. Of course you could have a dedicated DMA chip for memory to memory copies so it could be possible to unload the CPU from that task.
Of course when you are talking about things like netbooks the savings in cost may make any performance issues moot.
But I am not a hardware guy just a programmer.

Re:...Or an arms race

[idontgno]

Here are a couple of slot-mount SDD mounting frames.

But the idea of integrating a few score gigabytes of flash, with the appropriate controller and tied to a dedicated Southbridge SATA channel... brilliant! I don't know if anyone will go for it, though, except in the case of super-compact motherboards. Since losing enough flash cells means losing the entire integrated "hard drive", and therefore the full function of the motherboard. (Most motherboards don't have expendable components, like flash memory. Taiwanese knockoff capacitors notwithstanding.)

Re:...Or an arms race

[Totenglocke]

I think HDD will continue to stay enough ahead of SSD in raw capacity that it will stay relevant for a long time. When SSD is affordable at 200 GB then HDD will already be affordable at 2 TB, etc.

Then you install your OS and programs on the SSD and keep your files (music, videos, documents) on the HDD - not too different from what I do now where I have my OS and programs on my HDD and my files on an external HDD (I know it's not faster, but I have multiple computers and limited hard drive space on some, so it's easier to just move the external drive, plus that way if I need to reformat it goes a lot faster).

Re:...Or an arms race

[Richy_T]

And a waste of a PCIe slot.

Re:...Or an arms race

[ColdWetDog]

A lot of things can change in 10 years. Just think back to the year 2000: computers only ran at 2-3 GHz and Linux was just getting into the mainstream!

Seems like a slow decade, I guess. Computers still run around 2-3 GHz and Linux is still just getting into the mainstream.

Que no haya novedad. (May no new thing arise).

Re:...Or an arms race

[maxume]

The could be selling those at a loss, especially if you throw in the entire supply chain (that is, it may be 'cheaper' to sell the drives at $40 than it is to throw them away, even though they cost $60).

I have no idea if this is the case or not, I'm just pointing out that the retail price doesn't necessarily reflect cost to manufacture the item. Looking closer, Seagate does list the 160 GB drives on their website:

http://www.seagate.com/www/en-us/products/desktops/barracuda_hard_drives/

(The 160 GB should appear in the dropdown selector).

So presumably the drives are indeed wholesaling for less than $40.

Re:...Or an arms race

[Hadlock]

I don't know if anyone will go for it, though, except in the case of super-compact motherboards. Since losing enough flash cells means losing the entire integrated "hard drive", and therefore the full function of the motherboard.

Although I think that's where the market is headed. My mother had a perfectly good, roughly 8 year old, 3ghz pentium 4 that had performed reliably and is still powerful enough for what she needs to do. But the headache of removing 6 years worth of viruses, spyware and god knows what else just wasn't worth it. Between upgrading to 2gb of ram, wireless networking, a second video card (for dual monitors), a copy of windows 7 and replacing the buzzing fans and grinding hard drive, it was about the same cost to just replace it with a $300 nettop that bolts right to the back of her monitor that's arguably less powerful than her old computer, but I can lock it down and comes with free MS anti-virus software, and I don't have to worry about legacy support for XP in 2-3 years.
 
My point is, at this point, most people can replace the entire computer when the old hard drive fails for less than the cost to buy a new hard drive and pay for Geek Squad to install it for them. Computers bigger than a shoebox will cease to exist outside of the gaming community and some professions in another 2-3 years. Hell, we might even see some shoebox sized computers with horizontal risers with 2x pci-e slots for gamers, too. A dual core atom computer can play most computer games today, it's just the video card that's holding it back.

Re:...Or an arms race

[MrNemesis]

Agreed, and the nearline storage we implemented in my response to amorsen's post was along those lines - the "software" we wrote was about 250 lines of python and a couple of bash scripts for maintenance, with the rather awesome rsync+hardlinks doing the heavy lifting - I use a similar homebrew thing for backing up my home network, which implements (whole file) "deltas", date/time versioning and date pruning. I'd love to do the same thing at work with access to "proper" backup hardware + APIs (for things like taking consistent snapshots of DBs without dumping them to file) but the money for it isn't there.

Point is very much taken about tape reliability (although hard drives really don't travel well at all), we've had at least two that came back corrupt following a test DR. However, most of our contracts mandate tape in offsite escrow, and no beancounter wants to pay for both hard drive AND tape backup.

Re:...Or an arms race

[noidentity]

Right now you can buy a 1TB drive for right around $90. But you can not buy a 5ooGB drive for $45 or a 250GB drive for $22.50. There is a limit to how cheap you can make a harddrive.

Good point. Even a 1 MB hard drive must have the platter, head, servo, drive electronics, and enclosure. A 1 MB SSD just needs the controller chip and a tiny memory chip. It's easier to assemble too, just a couple of SMT components on a board.

BTW, consider using paragraphs.

Re:...Or an arms race

[mswhippingboy]

This sounds like the same argument I had with my boss 30 years ago. He wouldn't believe that microprocessors would one day obsolete the mainframes. He reasoned that no matter how powerful microprocessors got, mainframes would always stay well ahead of them. To my thinking, when it comes to electronics (both in speed and capacity), the distance between circuits begins to matter a lot when your working at the speed of light on a nanometer scale. There is abosolutely no way, in the long run, mechanical devices of any type will be able to stay ahead of purely electronic devices. Actually I'm surprised it's taken this long to get to this point. Costs will come down (I can pick up an 8GB thumbdrive for about $20 as opposed to the 8GB harddrive I paid $150 for about 10 years ago).
Spinning drives still have a few years left, but the writing is on the wall.

Re:...Or an arms race

[MrNemesis]

Figured you were a programmer (I'm an infrastructure guy so have only a murky view of the internal workings of software) - part of the question about the apps was why two different speeds of memory would be an issue - I figured it'd be bad practice for an app to assume that it'd always be able to map memory at 20GB/s or whatever, but I don't really know the ins and outs of low(ish)-level programming.

I figured it'd work in a similar way to file cache now - a map of blocks/tables is given to the OS to use for storage, and the kernel just alternates between flagging these blocks as in or out of memory as applications demand - effectively using memory as read/write-through cache. As soon as a DMA came in, the OS would read those blocks from SSD into memory transparently, and keep them there for as long as necessary. Similarly, a file object that's opened read/write in memory and has been static for X seconds would be transparently written back to the SSD, or immediately written back to disc as soon as the file handle is closed. I figure a small battery and/or capacitors would be able to handle keeping the data cached (or even written) in the event of sudden power loss to help prevent corruption, plus I'm pretty sure you could implement journaling as well.

But again, IANAP so please feel free to pick holes in my naive assumptions :)

Re:...Or an arms race

[profplump]

If you can afford to run 10Gbps to your off-site backup location it's either not far enough away to be useful or you're so rich that the shipping costs of tape vs. disk are irrelevant.

Re:...Or an arms race

[LWATCDR]

The problem with mapping mass storage to memory at the application level becomes one of locking and threads.
You will have a very difficult time with any threaded applications if you just do a map to memory. Locking goes from being a real pain to a total freaking nightmare.
In effect traditional disk io acts like a message passing system which is much simpler to deal with when doing threaded applications.
The real issue I think was a failure to communicate between you and m.dillion.
m.dillion was thinking of the programing model where you can actually open a file and it looks like a big honking block of memory. That model is really handy in a single threaded application but really starts to become a big mess when you are doing more than one thread.

Re:...Or an arms race

[SatanicPuppy]

I still have one from when they were 6 deep x 4 wide, and about 2 inches thick...A big bad 32mb drive.

Re:...Or an arms race

[Mad Merlin]

Sorry to burst your joke, but in 2000, CPUs had only just hit 1 GHz.

Re:...Or an arms race

[gwern]

Excess capacity summons its own uses.

You assume that your movies and videos cost only their storage space. But what if you switch to a finegrained snapshotting filesystem? That could use up a great deal of space depending on how much you download, modify, and delete.(With space no longer an issue, longevity becomes an issue. I'd rather lose half the space on an oversized hard disk if that means that 20 years from now I will still have my files.)

You assume you'll only store what sort of things you currently do. But what if you scan all your books to save space? High resolution book scans can easily consume hundreds of megabytes a piece. I was thinking of scanning an art book at best resolution, and calculated that its ~500 pages would use up around 20 gigabytes in TIFFs.

And what of new electronic stuff? Lifelogging (http://en.wikipedia.org/wiki/Lifelogging) is currently the stuff of experiments and early tentative ventures by the likes of Gordon Bell. But just yesterday the NY Times covered a small camcorder which uses 2GB for 72 minutes of video and costs ~$200; not useful, no, but useful in a few more years/cranks of Moore's law. Lifelogging is easily a gigabyte a day. That's peanuts, of course, and ever cheaper - but the point is that it will soon burst the seams of your 500GB drive.

And so on. Right now you think your needs are met. But there's always something more one could store, and needs change. Maybe, just maybe, you *are* right and 500GB is all you'll ever need; but for the rest of us, that's as sensible as Gates's mythical '540k is enough for anyone'.

Re:...Or an arms race

[zdzichu]

Well, we already have standard for onboard flash connector - it's just a matter of implementation in mainstream. Serious business already uses flash modules without SATA interposer. Just google for images of Oracle F5100 Array.

Re:...Or an arms race

[stewbacca]

However, for the average home user, "average" sized HDD are already far bigger than most people will ever need. The point of diminishing returns is upon us.

Re:...Or an arms race

[afidel]

No, it doesn't because an online backup is not the same as an offline backup. It only takes one pissed off sysadmin or one software snafu to blow up all your online backups, it would take the destruction of 3 separate facilities for all my offline backups to go away.

Re:...Or an arms race

[Gordo_1]

No, there won't be an arms race. HDD technology is pretty much at its limit. Spinning drive technology may allow for one or two more capacity doublings over the next decade (though I'm skeptical given how long we've been stuck in the 1-2 Terabyte range), but performance is pretty much maxed out and presents a substantial bottleneck for modern systems (think random reads/writes). Meanwhile, you can already purchase standard 2.5" half terabyte SSDs for less than $1500 (http://www.pricewatch.com/gallery/hard_removable_drives/ssd_512gb) and high end SSDs will be at a terabyte this year from multiple manufacturers.

I expect it will be less than 5 years before you can walk into Best Buy and plop down a couple hundred bucks for a multi-terabyte SSD for all your movies and media.

I don't even think HDDs have the features to replace tape for backup as they're too complex and fragile. Tape is appealing because it's inherently portable media, the media itself is very simple to store, reasonably rugged and will likely always offer a capacity premium over HDD.

Re:...Or an arms race

[GenP]

we'll need an OS-agnostic method of using PCIe cards as bootable block devices

Like pretending to be a single-channel SATA disk controller with one drive permanently attached?

Re:...Or an arms race

[denobug]

I am seeing a lot of discussion as far as cost/capacity, speed, tolerance rating. I am surprised no one has brought up the discussion of HDD and SSD in RAID configurations. I think one recent slashdot article discuss about the use of SSD in data centers and its limitations but so far this has not being part of this discussion.

Re:...Or an arms race

[MrNemesis]

Most likely as a new sort of controller hanging off the PCIe bus with a generic interface for plugging in flash cards of X gigabytes. Most of the ATA protocol is geared around spinning platters, and as such has a whole load of folderol that it and the OS needn't bother with for solid state media. KISS!

And in the event your post was intended as a jest, I'd rather see it implemented as a softraid driver of USB floppies, each connected to a 1.44MB flash chip ;)

Re:...Or an arms race

[Capt.DrumkenBum]

Everything old is new again.
Many many years ago I had a 20MB (Yes, Meg) hard drive that plugged into an ISA slot.

Re:...Or an arms race

[EvanED]

Sorry to burst your joke, but in 2000, CPUs had only just hit 1 GHz.

It's not so inaccurate... according to Wikipedia, November 2000 saw the release of 1.5 GHz CPUs, and 2 GHz P4s were out in August 2001. Even 3 GHz was out in Aug

Re:...Or an arms race

[gwjgwj]

I could see a netbook with a flash drive right on the motherboard.

It' s called eMMC and you can find it e.g in n900.

Re:...Or an arms race

[TheRaven64]

It would be possible, but horrible. For PCRAM or MRAM it makes sense, but Flash can not be written on a byte level (well, it can, but only once and then you have to erase the entire cell). You could make it read-only into the physical address space and then use another mechanism for writing, but that would be painful.

Re:...Or an arms race

[amorsen]

LTO-4 is a meagre 800GB per tape. 2.5 LTO tapes take up more room than a 2TB hard drive.

Re:...Or an arms race

[amorsen]

10Gbps is just a rented fiber away. As long as we're talking mere tens of km, that's fairly cheap. As soon as we're talking distances where you need active equipment in between it gets expensive, but I bet most tapes are kept within 50km of the tape drive used for writing them.

Re:...Or an arms race

[amorsen]

There's no reason that your pissed off sysadmin needs access to the systems at the other sites. Anyway, it's conceivable that some security policies will require the use of tapes. I just doubt it'll be common enough to keep the tape industry going. The real money is in tape robots and those are just as much online as the disk arrays.

Re:...Or an arms race

[adisakp]

No need for an arms race... right now they coexist quite well. Get a small SSD just big enough for your boot drive and a large HD for your data drive. It's the best of both worlds. They are already offering laptops with hybrids and the new SSD's have the potential to be very small so they could easily fit in most laptops with a second normal HD if they introduce a smaller SATA format.

Re:...Or an arms race

[evilWurst]

SSDs have fixed per-drive costs/requirements too. Chips to handle the SATA interface and internal wear leveling, for example. And the choice of memory chips is analogous to magnetic platter density. SSD makers spread the load over multiple memory chips, and spinning hard drive makers spread the load over 1-4 platters, but the makers of the chips and platters prefer to only churn out mass quantities of their one newest model. The investments for both factories are up front and then the marginal production cost per chip/platter is about the same, so why waste production capacity on the old model? That means, for both, that when capacity goes up, you may get a higher capacity drive for the same price, but you are unlikely to get an old-capacity drive for a lower price. There's a sweet spot around $75-$100 for hard drives and it hasn't changed much in ages, excepting discount selloffs of old stock when the new model's production has ramped up.

For example, checking newegg right now: cheapest 32 GB SATA SSD is $90, cheapest 16 GB SATA SSD is... $99. (Note that I'm comparing SSDs only, you'll have to skip by the expansion card drives). They declined to make cheap 16 GB MLC flash drives, instead doing SLC for those.

For comparison, if we hop over to the section on SD cards - which don't do anything fancy with drive controller chips in the card, they're pretty much just the memory chip in a plastic sleeve - we see the prices are much more closely related to capacity. 32 GB cards at $72, 16 GB at $32, 8 GB at $15.

Re:...Or an arms race

[bill_mcgonigle]

Hell, there's no reason why they couldn't just integrate a 20 or 40gb SSD right into the motherboard.

I opened up an new HP server the other day to find an SD card slot on the mobo, for 'embedded' booting.

Finding an SLC SD card in the 16GB range is a bitch, though. I've had so many MLC flash parts go bad-block on me ('wear-leveling', hahahah, I believed that 5 dead boot drives ago) that MLC's not really an option. Only one slot, no RAID.
 

Re:...Or an arms race

[bill_mcgonigle]

For PCRAM or MRAM it makes sense, but Flash can not be written on a byte level (well, it can, but only once and then you have to erase the entire cell)

Hrm, are there filesystems which fragment on purpose so you can gang your writes into blocks on flash and updates the pointers (by the block, of course) to locate? Seek shouldn't be a problem anymore.

Re:...Or an arms race

[LWATCDR]

Well the flash controller has to do that now. As a cost savings you could move that functionality to the PCs chipset or right into the CPU.
I think that is what they do now on smartphones and such.
Would it be ideal from a performance point of view? Maybe not.
But from a cost point of view? Maybe.

Re:...Or an arms race

[BabyBrumak]

The transfer speed is still too slow to SSD/HDD storage to be feasible. OS's already use page files on the drives to extend the memory space and improve performance. However, you could conceivably create a hybrid memory unit that uses the slower SSD along with high-speed RAM (similar to high-performance NAS systems) that would potentially add a huge performance boost.

Re:...Or an arms race

[LWATCDR]

I think people are not getting what I intend. You would still access the flash as a drive at the application level through a device driver. It would just be on the address buss and not on some external buss. You would still have ram based buffers and caching but would just use the CPU to deal with the flash instead of going through the IO controller to the SATA controller to the Flash controller and back again.

Re:...Or an arms race

[kextyn]

I strongly disagree with your statement about dual core Atoms. I wouldn't even want to use one with dual 5890s. While GPUs take up a lot of the responsibility in gaming the CPU is still very important. In one test I see on a hardware review site a 1.8GHz quad core Phenom barely reached 60FPS...at 1280x800 with medium settings in Far Cry 2.

Re:...Or an arms race

[Hadlock]

Oh, I meant the vestigial PCI (not PCI-e) slot on most full size ATX boards.

Re:...Or an arms race

[Hadlock]

Oh, I was talking about engines like the Source engine and Unreal 3. Counter Strike: Source will reputedly run at 15fps on a N270 (single core) netbook using the lousy GMA950. The Crysis engine is so lousy I don't think anyone would realistically try and run it on a netbook :)
 
Also: [geezer] You kids these days! Why, back in MY day, we ran Counter Strike 1.6 in software mode at 15fps, and we liked it! We were elated if we got anywhere near 30fps in our cutting edge games. You young whipper-snappers and your sixty frames per second...[/geezer]

Re:...Or an arms race

[Znork]

I think you underestimate just how much storage video can eat. I can easily see applications like PVR's that basically records all channels all the time and lets you simply pick what you wanted to watch post-broadcast. Something which could easily eat all the storage you could throw at it for a long time...

A petabyte isn't much these days; when my corporate co-workers lose perspective and think we have lots of storage as we're a big corporation with many petabytes, I point out that the many regular online computer shops where we buy consumer parts from have several times our total storage volume lying around in stock for next day delivery. Corporate storage may be expensive, but the volumes seem to grow far slower than consumer storage.

Re:...Or an arms race

[don_bear_wilkinson]

But to say that spinning disks will go away is kinda short sighted... You meant long-sighted. FTFY :)

Re:...Or an arms race

[owlstead]

I don't understand. As long as you still treat it like a hard disk drive from within the OS and not just memory, why not use direct CPU addressing? I might be missing the point, so please enlighten me.

Re:...Or an arms race

[mog007]

While only having a single core.

Re:...Or an arms race

[Millenniumman]

HDD is already affordable at 2TB, the best deal being about $150. 200GB SSDs are, on the other hand, $900. 7 cents per gigabyte or 4 dollars. They aren't remotely comparable. SSD has to come a loooong way to even compete, and it's not as if they have huge advantages over hard drives anyway.

Re:...Or an arms race

[timeOday]

Not quite. A shop near my house (Rome, Italy) has 320GB drives on sale at 35.4 EUR (roughly $48) including 20% sales tax

$50 is a lot of money when you're talking about $100 netbooks! 40 GB is a fine amount of storage for such a computer and SSDs in that capacity will probably be 1 cm^2, soldered to the motherboard, and cost $15 within 2 years.

Re:...Or an arms race

[Dahamma]

I think you underestimate just how much storage video can eat. I can easily see applications like PVR's that basically records all channels all the time and lets you simply pick what you wanted to watch post-broadcast.

No, I know how much it can use. 1080p H.264 is 5-10GB per hour these days. A cheap 2TB HDD can record 200-400 hours.

Maybe "all channels all the time" could be recorded in a cable headend, but not likely in the home, as with any current system you'd either need a tuner/demod for each channel/frequency (cable/satellite) or an absolutely enormous amount of bandwidth (IPTV). And that doesn't even consider the disk transfer rates, which would make individual huge storage devices useless for this task. It would be FAR easier and more cost-efficient to keep that massive amount of information in one place at the headend rather than in everyone's home, and then just send the single stream to be watched to the consumer. Which is what services like Comcast OnDemand are already starting to do...

And I'm not talking about crazy hobbyist projects, but mass-market applications for HDDs for the consumer. And in the corporate sense, I'm also not talking about thousands of desktop HDDs in a large corporate environment, but centralized storage where you basically want the most storage in the least rack space.

So, I still stand by my statement... for the average consumer, data usage has not kept up with HDD sizes. After it doubles a couple more times, there just won't be much of a consumer mass market for high-end HDDs. In the data center, it might go a bit further (but again big, slow HDDs start losing some of their usefulness when you can't get the data off them in a reasonable amount of time...)

Re:...Or an arms race

[BikeHelmet]

The parts cost on an HDD and SSD is pretty close, though. Including two drives would be a premium feature in larger laptops.

I also disagree with whoever said 40GB is enough. 40GB isn't enough - not if you have a modern OS, anyway. Even my parents use more than 40GB, and they don't play any games or do any video work. Just photos from their camera, and documents, and programs. Mind you, it did take them 2 years to fill it.

To overtake HDDs, SSDs also have to beat them in price - not just match them. They're not going to match them in space (and consumers love big numbers), so it has to come down to price. Right now a low end 160GB drive probably costs OEMs about $30 or less per unit. While I could see an OEM replacing that drive with an SSD, I doubt they'll do it if it adds a lot to the cost, pushing it into the premium laptop/netbook range. Assuming $80 OEM pricing on an SSD... $50 is a big difference. Would you buy a $350 netbook if a $299 netbook was right next to it? Add to that that 40GB isn't enough space for a "Premium" model, and you have a "Pass" from OEMs, for at least one more generation.

But when Intel can provide 60GB/80GB SSDs to OEMs for ~$0.80/GB, that'll be the flipping point.

Re:...Or an arms race

[CAIMLAS]

At some point SSDs in the 120Gb range will be cheaper than spinning platters. It is probably close right now.

Where can you find a 120G SSD for under $50? You can get an iPod with a 120Gb drive for $20, but an 8Gb SSD is still $60.

Re:...Or an arms race

[Anonymous Coward]

This is weird... after reading your post, I just took a look at my local computer retailer's website, it just happend to be added a few days ago (sorry, German, but obvious)

http://www.kmelektronik.de/shop/index.php?show=product_info&ArtNr=22670

Re:...Or an arms race

[Hadlock]

Cooool! Now the next question is, can you install Win7 on 32gb?

Re:...Or an arms race

[Eivind]

Why ? We've got a fibrebundle (around a dozen fibres), it's not owned by us, but we can rent a single fibre across town for around $500/month, yes that's expensive, but it's not out-of-range for most companies, not even smallish companies.

Not that we need 10Gbps for offsite-backup, we currently do it over 100Mbps, and that link isn't on the average even 5% full.

Re:...Or an arms race

[RichiH]

Yes.

Hopefully, MRAM or similar will do away with all of them, and RAM, at some point.

Re:...Or an arms race

[TheRaven64]

While the write is in the queue, your memory mapped flash would be in an inconsistent state. This would be an absolute nightmare to program for. You'd effectively have to buffer everything with a RAM cache, as you do today, but because it is mapped directly into RAM you'd have to use the CPU for copying the data instead of the DMA controller, making everything much slower.

Re:...Or an arms race

[Drethon]

So MRAM is the new SSD? Sorry, just had to...

Re:...Or an arms race

[LWATCDR]

The only problem I see could eating up too address buss bandwidth servicing the slower Flash. Even that may not really be an issue except that CPUs are not intergreating DRAM controllers on the core so they may not know how to talk directly to Flash.
Things are no as simple as they where.

Re:...Or an arms race

[LWATCDR]

Couldn't you have the DMA controller handle the write from the ram buffer to the flash?
Yes I know that it probably would never be the fastest solution but It might be the cheapest and PCs right now seem to be fast enough for most folks.

Re:...Or an arms race

[TheRaven64]

You're missing the point. With a typical OS implementation, you copy 4KB from disk into a RAM cache by DMA. Then you let processes modify this. Then you write it back by DMA. The version in RAM is always consistent and the FS layer synchronizes this with the on-disk version.

With memory mapped flash, because you can't write anything smaller than a cell, you'd still need this buffer in RAM because, unless you wanted a read-only file you'd need to buffer writes and you'd need to read from the buffer otherwise you write then read and get the old version back (sometimes - fun bug to find).

For copying between the RAM cache and the flash, you want to use DMA or you tie up the CPU during the operation. If you're using DMA to move data to and from the flash, what benefit do you get from having it memory mapped? Because flash is significantly slower than DRAM, you probably don't even want to use the mapped version for read-only files. Embedded platforms do for a few things because it saves RAM, but this only makes sense where RAM is scarce (and they typically update the page tables to use a copy in RAM after boot when possible anyway).

You're adding an entire layer of complexity for no benefit. Flash is intrinsically a block device. Plugging it in via a block device interface makes more sense than trying to make it pretend to be a byte-addressable device and then having extra logic at both ends to work around the fact that it isn't one.

Re:...Or an arms race

[owlstead]

Very good points, we'll probably have to rely on the chipset to solve those problems - the trick is keeping the latency to a minimum.

I'm not so sure SATA is the way to go for addressing Flash (or hopefully, phase change memory) in the long haul. PCI already makes more sense, initially manufacturers could just emulate the SATA part in software or hardware if required - it's a bit tricky for boot-up using a standard BIOS though.

Re:...Or an arms race

[Just Some Guy]

Sorry to burst your joke, but in 2000, CPUs had only just hit 1 GHz.

Moore called and agrees: 2 GHz was a year later. He also agrees that a 1-part-in-2000 rounding error makes the joke unfunny, and that girls at parties are surely awed by your devotion to technical accuracy.

Re:...Or an arms race

[LWATCDR]

You save the cost of the flash ram controller and reduce the part count. Probably not by much but but depending on the read speed of the flash a read could be faster since it would just be a DMA on the address buss.
Mainly just to shave pennies for things like netbooks.

Re:...Or an arms race

[petermgreen]

One problem I see with having your backups on media that remains powered is if there is a seriously bad power surge (e.g. a lightning strike very close by) then anything connected to either the mains or a wired network is likely to be fried.

Re:...Or an arms race

[petermgreen]

I doubt it for a few reasons.

Firstly the bandwidth just isn't there in many places. Even areas with supposedly high speed connections often have very limited upstream speeds.

Secondly there is reliability, home and small buisness internet connections aren't always very reliable and there is usually no service level agreement on them at all.

Thirdly there is mobility, mobile broadband is even less fast and reliable than fixed line broadband.

fourthly there is security, would you really trust confedential buisness data to a cloud provider who likely has far more exposure than you do.

We don't even have the computers in our offices running diskless from local servers let alone running diskless across the internet.

Re:...Or an arms race

[petermgreen]

When 3.5" go obsolete once and for all the 2.5" drives will stop costing a premium
If you don't want all that much capacity and don't care about speed then the 2.5 inch drive is already sometimes cheaper. For example looking at newegg the cheapest 160GB laptop drive ( http://www.newegg.com/Product/Product.aspx?Item=N82E16822148443 ) is $39.99 with free shipping while the cheapest 160GB desktop drive ( http://www.newegg.com/Product/Product.aspx?Item=N82E16822148511 ) is $37.99 plus 6.25 shipping.

For capacities too large to fit on one side of one 2.5 inch platter I suspect making the platter 3.5 inch will always be cheaper than adding a second read/write head.

Re:...Or an arms race

[Jay L]

Sure, but with network-based backups, your media aren't all in the same place; they're in data centers or friends' houses or floating platforms around the world.

Re:...Or an arms race

[hitmark]

if only, tape drives are priced for corporate use, not home.

Re:...Or an arms race

[hitmark]

heh, why not go for software rendered quake 1? was some impressive mods made back then, unlike today when everything looks like bunnyhopping CS on caffeine and sugar (or maybe fructose?).

Re:In 5 years

[Anonymous Coward]

In 5 years, people will still be maintaining COBOL systems.

Price isn't everything

[Angst Badger]

Price is only the first hurdle for SSDs. There's also the issue of reliability, and reports from the field suggest that SSD reliability is highly variable, and in no case as good over the long term as hard drives. That will probably change in time, but they're not there yet.

Re:Price isn't everything

[Anonymous Coward]

SSD reliability is highly variable, and in no case as good over the long term as hard drives

SSDs are much like anything else, you get what you pay for.

Buy the cheapest hard drive you can find; it won't last 5 years. Same for CD-R media and SSDs.

It's a pretty safe bet that under normal workloads, a good SSD will outlive just about any HDD.

Re:Price isn't everything

[Silentknyght]

SSD reliability is highly variable, and in no case as good over the long term as hard drives

SSDs are much like anything else, you get what you pay for.

Buy the cheapest hard drive you can find; it won't last 5 years. Same for CD-R media and SSDs.

It's a pretty safe bet that under normal workloads, a good SSD will outlive just about any HDD.

I'm going to have to agree, especially considering I've just recently suffered a premature HDD failure. If you read the customer specs for any of the larger capacity platter drives, you really notice the failure rates: failures at the 3-6 month mark should be one in 100,000, not a one in 100 (or less).

Re:Price isn't everything

[amorsen]

Where did you find reports from the field? All I've seen are lab studies and guesses.

Re:Price isn't everything

[Lunix Nutcase]

If you read the customer specs for any of the larger capacity platter drives, you really notice the failure rates: failures at the 3-6 month mark should be one in 100,000, not a one in 100 (or less).

And where are you getting statistics that are showing that one in 100 hard drives are failing with in 3-6 months? Oh wait, you're extrapolating based off a biased sample.

Re:Price isn't everything

[Lord Ender]

SSDs from Intel, SuperTalent, OCZ, etc. are both fast and reliable.

Problems happen when you try to boot from thumb drive or otherwise misuse flash disks. But those disks which are designed by reputable companies to act as traditional hard drives are certainly more reliable than mechanical disks.

Re:Price isn't everything

[Nemyst]

At least SSDs tend to fail over time, meaning you only lose a fraction of your data (if any) unless you're being REALLY careless. HDDs have a tendency to just explode, leaving you with nothing unless you've done backups frequently. That's not counting advantages of speed, power consumption, size factor, heat generation, reliability potential (no moving parts means that there eventually WILL be a consensus that SSDs are more reliable than HDDs) and more.

I also haven't seen any evidence to say SSDs would be less reliable than HDDs. Remember it's easy enough to say that they "only" last half of what they could, but if that "only" is 5 years, then it's also the average HDD's lifetime.

Re:Price isn't everything

[Rockoon]

How about an unbiased sample? An extensive sample, even?

Google, being a very big consumer of HD's, has published such data.

3% die in the first 3 months, another 4% die by the end of the 1st year.

8% more die in year 2.

Maybe you have missed the fact that these monster capacity drives actually suck in the reliability department? yeah.. 3 out of 100 convert themselves into worthless crap within 90 days.

Re:Price isn't everything

[j-turkey]

obligatory link to Google study/data.

Re:Price isn't everything

[Anonymous Coward]

Where did you see these reports from the field? For one thing, when an SSD runs out of write cycles and fails, the data doesn't just vanish. It becomes a read-only device. So a failure isn't catastrophic. Recovery consists of shutting down, dropping in another device, copying data over, then rebooting again. That alone is a HUGE improvement over hard drives.

HDD reliability itself is highly variable. I have working 100-meg drives from the early 1990s (I have no idea why I've kept them except as curiosity pieces). Then I've had drives fail just days out of warranty. And the last conventional HDD I bought lasted long enough to install Windows and use it for a couple of days, then it failed spectacularly.

And if I write to an HDD and put it on a shelf, when I come back to it, will it still work? The data will almost assuredly be on the platters, but will the motor spin up? Will the heads stick? I know if I write to flash memory, I can set it aside for years and years and it will still work, because we've been using flash memory for storage for at least the last 15 years. Will my Pentium-75 motherboard from 1995 go into the BIOS if I apply power and hit DEL? Yes. So there's proof of flash memory's longevity.

I switched over to OCZ Vertex SSDs in my home PCs about a year ago and couldn't be happier. They've been highly reliable, and not just fast--they're stupid fast.

Re:Price isn't everything

[rrohbeck]

I have yet to see one SSD failure (and I work with them all the time, mostly Intel X25s.) As opposed to several HD failures - either completely or bad sectors.
The 160GB hard drive in my eeePC just failed to read an MP3 file. I guess I'll replace it with an SSD soon.

OTOH, USB sticks and SDHC cards do fail all the time. I guess they have less redundancy and error correction. Just for fun, I ran Ubuntu off an 8GB SDHC card in the eeePC for a while. The performance was acceptable with a fast card, but there were hard errors after a while.

Re:Price isn't everything

[bill_mcgonigle]

Where did you find reports from the field? All I've seen are lab studies and guesses.

Newegg reviews. I'm nearly serious.

Re:Price isn't everything

[owlstead]

In a server environment...where they are accessed continuously.

Re:Price isn't everything

[CAIMLAS]

Unfortunately for hard disks, their reliability has also been horribly variable in the past year or so. We've had an obscene number of 6-month-or-newer disks fail recently, and it's been a huge pain in the ass.

Excellent

[__aaelyr464]

This is great for the average joe computer geek like myself... I don't need some high end SSD drive for a laptop I use for basic work. But the idea that I could get an affordable SSD for it and still reap many of the benefits is great to hear.

Re:Excellent

[Vectormatic]

agreed, i'm seriously eyeing an SSD, and for my linux uses 40 gb should do just fine

now if only some mobo maker would produce a nice socket am3 mini-itx mobo with dual dimm slots and a pcie 16x slot (and amd 785g chipset please!) i could build my perfect little linux box

Careful on Your Terminology There

[eldavojohn]

SSD Price Drops Signaling End of Spinning Media?

Blu Ray and CDs are still "spinning media" aren't they? I think I've seen many holographic storage disc products (touted to be THE FUTURE) that were spinning as well. I agree that our mechanical media may be just atop the apex or turning point but our non-mechanical disc based media is most likely set to be a some form of spinning disc for at least a few years longer. If the article thinks that movies and albums will switch to SSD based distribution, I just don't see it happening real soon or even now.

Re:Careful on Your Terminology There

[idontgno]

Blu Ray and CDs are still "spinning media" aren't they?

To be replaced with network-accessed or network-streamed material. Read-only rotary optical media will be a "way back" story our children will tell our grandchildren. (In other words, my 4-year-old daughter will tell HER 4-year old daughter "I used to watch Dora the Explorer on DVDs.")

Re:Careful on Your Terminology There

[LWATCDR]

"If the article thinks that movies and albums will switch to SSD based distribution, I just don't see it happening real soon or even now."
Oh no you are right. No one will ever download or stream music and or movies. And nobody will ever sell music on SD cards.
Actually I think we will see HDDs for some years yet but I actually think CDs and DVDs may have an even shorter lifespan.

Re:Careful on Your Terminology There

[Hadlock]

I'd love to see the next version of windows released on a USB thumb drive. I own one external usb DVD player for OS installs and that's it. It feels very 1995 to still be installing my OS from optical media.
 
As Netflix expands their streaming movie selection and increases the video quality of their streamed movies, I think Blu Ray is going to die, or at least wither quite a bit in first world countries. Sure the audio/videophiles and people who like to collect things (like movies) will still buy their favorite titles and Blu Ray will continue to live on, but most people are content to skip the physical disc if they can get streaming videos ala carte at a flat rate. In third world countries VCD and DVD rental places are quite popular where broadband exists, but it still very slow, so I can see Blu Ray flourishing there once the rates for duplication and players comes down by at least 70%. I don't think we live in a world where silicon memory chips and their controllers will ever be cheaper than $10 or so, which is significantly higher than the cost of a piece of plastic with some foil embedded in it.

Re:Careful on Your Terminology There

[blair1q]

CD is already dying. Blu-ray is about to get its ass whupped once someone starts distributing movies on micro-SD, or the market shifts to movie-downloading services. A 4-GB micro-SD can hold 1080p movies. Overnight download requires no physical distribution at all.

We're close to the emptying of our computers and media centers of all their machines (except the nifty motorized volume knobs; those will never die). Even muffin fans are becoming less necessary.

Re:Careful on Your Terminology There

[amorsen]

CD's are 800MB. You could easily replace them with SD cards without significantly affecting what the music industry earns.

But why would you replace one physical medium with another, when you can just switch to online distribution?

Re:Careful on Your Terminology There

[0123456]

Overnight download requires no physical distribution at all.

Because everyone wants to wait until the next day every time they want to watch a movie.

And every time the network does become fast enough to watch that movie in real-time, the next standard comes along which makes it too slow; 4k images and 3D movies being the next two that the movie industry is likely to be pushing on consumers.

Re:Careful on Your Terminology There

[Jenming]

Streaming video is great for a computer screen or an SD TV. But streaming enough pixels to fill a 1080p screen requires a lot of bandwidth. Blu-ray spec is 54Mbps at the low end.

While some people may be lucky enough to have that in download speed, most don't and upload speeds like that are a long way off.

Re:Careful on Your Terminology There

[fermion]

I see spinning media, blu ray, CD, HD, as being used primarily for backup. Already the external hard drives I have are for backup, while anything I use regularly is solid state. Probably the small computers I buy will be solid state. I can see the time in five years when anything but a desktop computer will be solid state. The DVD player is about the least reliable and most power hungry part of on my laptop, so I can see wanting that to go as well.

Re:Careful on Your Terminology There

[MobyDisk]

Yes, DVDs spin 2000 RPM, while HDD's spin between 5400RPM and 15000RPM. Big difference in power and heat. Plus, hard drives spin fragile magnetic disks on exotic fluid bearings with a physical head micrometers away. They are a lot more susceptible to damage.

Although this still misses a big point: DVDs and CDs are going away too.

Re:Careful on Your Terminology There

[evilviper]

Blu Ray and CDs are still "spinning media" aren't they? I think I've seen many holographic storage disc products (touted to be THE FUTURE) that were spinning as well.

CDs are tiny... ~12 full CDs will fit on a $30 USB thumb drive. Blu-Ray isn't all that big, either... ~40 Blu-Ray movies on a $100 HDD?

Optical media will succeed only if densities can continue to increase, all the while the pressing technology remains fairly simple. As soon as Disc+ yields / speeds are low enough that writing data to Flash is faster, discs will go away for good... The ability to stamp out discs at high speeds and low costs is a great benefit, but the drawbacks will kill the medium as soon as those benefits aren't so huge anymore... For example, if the number of layers on a disc has to climb much more than 2 to keep up with desired capacities, expect prices to rise, substantially.

And holographic discs are the ultimate in vaporware... Slashdot has been having stories on multi-terabyte HVDs since '99. They come with a massively expensive product they swear will be dirt cheap in a month, they get a bit of funding, then they fizzle out...

Re:Careful on Your Terminology There

[Neil Hodges]

Yeah, if you have Internet access. Doesn't work so well when you don't.

Yes, it may be hard to believe, but there still exist places where Internet access isn't practical (generally due to high cost).

Re:Careful on Your Terminology There

[Idiomatick]

Once copyright issues get worked out CDs and the like will die at an alarming rate. Nearly everyone will have access to the net so buying things on disk will become pointless.

Re:Careful on Your Terminology There

[afidel]

Blu-ray spec is 54Mbps at the low end.

Actually that's the MAX datarate and the max video+audio rate is even lower at 48Mbps, that's high for the US but already here for many other countries.

Re:Careful on Your Terminology There

[TheRaven64]

And every time the network does become fast enough to watch that movie in real-time, the next standard comes along which makes it too slow

And a smaller subset of the target market cares about the quality. I can stream 720p from iPlayer now. In a couple of years, I'll be able to stream 1080p. Will I be able to tell the difference between that and 4K? Maybe. Will I care enough to trade the convenience of having a stream start a few seconds after I want to watch it for the increase in quality? Hell no.

Re:Careful on Your Terminology There

[rrohbeck]

Blu Ray and CDs are still "spinning media" aren't they?

Yes but nobody in a sane state of mind uses them for storage or backup.

Re:Careful on Your Terminology There

[bill_mcgonigle]

Yes, it may be hard to believe, but there still exist places where Internet access isn't practical (generally due to high cost).

I'm in the US. Something like 21% of the population can't or doesn't want to access the Internet.

Here we have discount stores which sell the spinning plastic in very large volumes for $6.

Re:Careful on Your Terminology There

[Confuse Ed]

Floppy disk drives (0.1 to 3MB), ZIP drives (100 to 750MB) have been totally replaced now by solid-state drives (in the form of USB memory sticks) - and re-writable CDs and re-writable DVDs are pretty much dead except for specialist cases (maybe medium-term archiving or transfering data to other devices like DVD-players or CD music players ; but for data-transfer an 8GB USB stick is far more convenient than burning a DVD).

Read-Only media for mass distribution is another matter entirely : physically stamping out the data in disks that can be spun past the reading mechanism like in a CD or DVD is unlikely to be superceded by something without moving parts for a while - although it is not unfeasible to imagine something like a CD but with the media stationary and the reader moving / directing its reading mechanism / beam accross it in 2 (or 3...) dimensions its a lot simpler / cheaper currently just to spin the disk around to provide one dimension of scanning (Simplifying the complicated / expensive movement of the heads to just one dimension while still allowing access to the whole 2D surface)

Re:Careful on Your Terminology There

[Hurricane78]

Blu Ray and CDs are still "spinning media" aren't they?

Yes. And USB sticks still have already put an end to the latter, and will do so to the former too, soon.

Re:Careful on Your Terminology There

[Jenming]

How fast can you read/write data on a Blu-ray disc?

According to the Blu-ray Disc specification, 1x speed is defined as 36Mbps. However, as BD-ROM movies will require a 54Mbps data transfer rate the minimum speed we're expecting to see is 2x (72Mbps). Blu-ray also has the potential for much higher speeds, as a result of the larger numerical aperture (NA) adopted by Blu-ray Disc. The large NA value effectively means that Blu-ray will require less recording power and lower disc rotation speed than DVD and HD-DVD to achieve the same data transfer rate. While the media itself limited the recording speed in the past, the only limiting factor for Blu-ray is the capacity of the hardware. If we assume a maximum disc rotation speed of 10,000 RPM, then 12x at the outer diameter should be possible (about 400Mbps). This is why the Blu-ray Disc Association (BDA) already has plans to raise the speed to 8x (288Mbps) or more in the future.

http://www.blu-ray.com/faq/

Re:Careful on Your Terminology There

[afidel]

BD Video movies have a maximum data transfer rate of 54 Mbit/s, a maximum AV bitrate of 48 Mbit/s (for both audio and video data), and a maximum video bit rate of 40 Mbit/s.

http://en.wikipedia.org/wiki/Blu-ray_Disc#Bit_rate

It actually points to the source of the information which actually points to the specification =)

Re:Careful on Your Terminology There

[moonbender]

New technology tends to take a while to trickle down to the third world.

Re:Careful on Your Terminology There

[moonbender]

I'd love to see the next version of windows released on a USB thumb drive. I own one external usb DVD player for OS installs and that's it. It feels very 1995 to still be installing my OS from optical media.

It is.

(FWIW installing Linux from USB is, if anything, even more straightforward.)

Re:Careful on Your Terminology There

[petermgreen]

CD's are 800MB. You could easily replace them with SD cards without significantly affecting what the music industry earns.
I dunno what the real price of SD cards of that capacity is, I see most places selling them 512mb SD cards (a bit smaller than a CD but you could easily use some kind of compression on there to make up for that) are selling them at prices in the $5 range though a few are selling them a lot cheaper (most likely surplus stock), lets assume the bulk price is a quarter of that or $1.25 plus you have the cost of duplicating, labeling, boxing etc,
so say $2 per copy.

CD duplication in bulk seems to be about a $1 in medium quantity including cases and inserts http://www.discmakers.com/products/CD100S.asp , I suspect they get much cheaper in huge bulk.

Given that many CDs retail for less than $10 I suspect an extra dollar would be a pretty big chunk of the profit.

Child pornographers.

[MindlessAutomata]

Hard drives and very powerful magnets will still be the #1 choice for child pornographers.

Re:Child pornographers.

[floppyraid]

That is a really persistent myth (that magnets will erase/corrupt data on a modern hard disk drive).

Inside of all harddrives for the last 10 or so years are multiple, very powerful neodymium iron boron magnets that move the actuator arm over the surface of the discs. If magnets outside of your drive would erase data, then surely these intensely powerful magnets inside would do the same, no?

The most conclusive testing I've seen done on this was several years ago. A guy had stacks of dead hard drives, and he decided to harvest the magnets from them. He had a stack of 50+ very powerful NIB magnets. He then took a working HDD, full to capacity, and covered the entire hard drive in them- front and back, with layer upon layer of magnets. Then he set the drive in a desk drawer for a few weeks, after which he plugged the drive up, and all of his data was still completely intact. Not 1 file was corrupted in any way.

Now, if you put a .40 or .45-caliber round through a platter, you can be certain the data is unrecoverable. Last time I checked, HDD platters are made out of some sort of silicon composite, so a bullet should shatter the entire plater (or at least half of it) into tiny fragments.

Re:Child pornographers.

[Steauengeglase]

I'm sure they are already on top of that. I'd imagine a jumper cable would make a great charge pump. Now we just need to convince them that it only works in a flooded room while wearing metal shoes; then I'd say the circuit is complete.

Re:Child pornographers.

[stonewallred]

Nah, just as today. Deep basements are the answer to their storage needs.

Re:Child pornographers.

[Vectormatic]

CMOS chips and high voltage work pretty well for data destruction too

Re:Child pornographers.

[Hadlock]

Back in middle school me and my buddy wanted to try out linux but didn't want to wait to format* the drive so we stuck the magnet out of the base of a magnet-mount shop lamp (10 lb "capacity", about 5" in diameter). To our surprise, not only did we corrupt the drive data, but the computer wouldn't recognize the drive, either.
 
*I am aware now that there's more involved to formatting a drive

Re:Child pornographers.

[TheThiefMaster]

These days you can "quick format" a drive, which only alters the file table and so on, it doesn't zero the whole disk. Zeroing the disk takes so long and is of only dubious benefit.

Re:Child pornographers.

[Amouth]

unrecoverable is relative to how much money you throw at getting it back - if all you do is break the platter up then the data is still there and using a scanning electron microscope you could retrieve the data off the pieces and then put it together.. not cheap but it would work.

i'd have to find the article but i remember reading that OnTrack managed to get 99.99% of the data off the drives on the most recent shuttle explosion. i'm willing to bet if they can get data off drives that went through reentry that they can get it off a drive that was shot.

Re:Child pornographers.

[Hadlock]

I'm reasonably sure quick format wasn't an option back in Windows 3.1 waaay back in 1993, but I will go ahead and forward that to my past self... Does the pop3 standard handle temporal shifts ok? Maybe I should send it pop2 just in case the servers from back then don't understand pop3 :)

Re:Child pornographers.

[bored]

None of the modern IDE's let you "zero" the disk, or even fully format it. Without the embedded servo tracks the head can't even know if its staying on track.

Re:Child pornographers.

[floppyraid]

Yeah, that is true. The only way to be completely sure would be to do some serious damage to the platters and scatter their remains in various places.

But also, with modern drives it would probably be much harder to recover data after a major catastrophe like that. I remember reading that article about them eventually finding the hard drive from the shuttle- it was an old Seagate made way before 2000. I was pretty amazed that it survived, but I do believe that the density difference in storage nowdays would have made it very very difficult if not impossible to recover any large amount of the data off a newer drive under similar circumstances.

Modern drives are not made a durable as those older drives were, and, they are packed far more dense-- the first 1TB drives were using 5 platters, nowdays I believe the latest Samsung Spinpoint F3 only uses 2 platters -- that's 500GB per platter-- Thats a huge jump. I am almost positive that the Seagate drive from the shuttle disaster was less 1GB, probably 500MB, and probably had at least 3 platters- the difference in density is massive.

Nevertheless that is amazing that they were able to recover the data. I think it would be rad (albeit probably very tedious) to work in a clean room doing serious data recovery. I wonder how much that data recovery ended up costing them? Almost certainly over 10k

Re:Child pornographers.

[TheThiefMaster]

Apparently POP3 is from 1988. A 1993 RFC talks about it as a "draft" standard, but also mentions "the optional APOP facility (which is in interoperable, operational use in at least three implementations)" suggesting that despite being a draft standard it was already in wide use.

So you should be fine to use pop3.

Re:Child pornographers.

[StikyPad]

Whew. Thank God you were here to discredit an entire industry, along with NSA/DOD approved methods for the declassification (aka destruction) of classified government media.

Now if you'll just update this Wikipedia article, we can all sleep soundly, knowing that our magnetic media is indeed impervious to magnetic fields.

Re:Child pornographers.

[Idiomatick]

Are you giving an endorsement then?

Re:Child pornographers.

[TheRaven64]

format /q was an option back in the DOS days, and did the quickformat thing. I think it was introduced with MS DOS 5, but it might have been DOS 6. To install a different OS, however, you don't need to format the drive at all. Just delete the partition table and start again. This involves editing a 512 byte block (which the install tool does for you, and did even in the DOS days). There is no need to erase the existing data, just write over it.

Re:Child pornographers.

[afidel]

Exactly, even a decade ago the huge degausser's the US Navy uses on their ships weren't powerful enough to wipe a HDD, ever since HDD's started using the GMR effect in 1997 in fact.

Re:Child pornographers.

[ImprovOmega]

I would say unrecoverable it relative to how much money you throw at destroying it.

Degaussers are pretty good at leaving data unreadable.

A pit of molten white hot steel will pretty thoroughly destroy all information on a drive.

Launching the disk platters into the sun would be practically certain to eradicate all data for all time.

However, for the ultra-paranoid, finding a nearby black hole and dumping it in there is the only way to be 100% theoretically certain that all data is destroyed.

Black holes: nature's ultimate shredder.

Re:Child pornographers.

[Anonymous Coward]

So how do you know the data was corrupted if the computer wouldn't recognize the drive??

Re:Child pornographers.

[Rockoon]

I think it was more like DOS 3

Nikon F6 and FM10

[BadAnalogyGuy]

There are only two advantages SSD has over spinning media at this time: Access speed and Durability. Storage space is still not up to par, and cost is definitely a weak point. However, technology progresses and we're hitting the limits of the current hard disk technology. SSD technology is definitely the future of most personal storage.

But it won't replace it in all areas. There are still "obsolete" technologies in widespread use due to technical superiority over perceived convenience. No one is going to say digital cameras are lousy, but compared to film, they are simply outmatched. Where is Velvia for digital? Where is Kodachrome? These films have no equal in the digital world except as poorly implemented filters in Photoshop.

Spinning media is going to be with us for a while, and I expect, like film, that eventually prices will go back up and this technology will be a specialty market targeted at high-end users and professionals.

Re:Nikon F6 and FM10

[LikwidCirkel]

Except, chemical film stores the data in a different way than digital cameras, and there is a significant difference between the two because of this. With hard drives, the data is stored in the same way - as bits - regardless if it is spinning or not. I don't see how the details of _how_ bits are stored matters to the high-end user. He or she will just want whatever solution stores the bits the fastest and most reliably.

Re:Nikon F6 and FM10

[natehoy]

Durability? Umm, yes and no.

You can beat a properly-built and operating SSD with a stick and there's a very low likelihood of damage, while the spinning hard drive will almost die an immediate and horrible death. And as long as the controller holds up, whatever data is on the SSD will almost certainly be OK on it. So, physically, a well-built SSD is a tough little rascal.

But the media contained in it has a lifespan of somewhere on the order of five years of relatively ordinary use. The spindles might not fail, since it has none, but you'll find slowly diminishing capacity and performance near the end of the lifespan of the drive. Compare this to spinning hard drives, where I have a 20GB drive I bought when 20GB drives were new (late 90s) that still performs as well as the day I purchased it. I have SMART monitoring on and it keeps coming up all-green. It's been in continuous use for well over ten years. I use it only for an OS drive and don't keep any data on it any more because one of these days it's just going to die, but I'd be nearing end-of-life on my second SSD drive by now.

Now, there is one advantage of SSDs. They fail earlier, but the failure is usually in the form of loss of capacity and performance. Drive failures are rarely as dramatic as a spindle lockup or head crash, so you can usually get your data off the drive while it still works fine. Data-loss failures do occur, but they are (AFAIK) somewhat rare.

Re:Nikon F6 and FM10

[natehoy]

Whoops, I said "SSD drive", didn't I? Did I really type that?

And me a Grammar Nazi. Have at it, folks. I have it coming. ;)

Re:Nikon F6 and FM10

[Anonymous Coward]

Oh please, Get over your aesthetic snobbery, luddite.

Velvia and Kodachrome produced exposure styles that are fairly trivial to reproduce in both Gimp and Photoshop- Film locks you in to an entire set of aesthetic preferences by brand, Digital gives you a raw capture and leaves it to the photographer to apply the aesthetics after the fact (or Batch Apply, if you find a style you like).

If anything, these days, the 'consumer aesthetic style' of film is being replaced by the camera bodies themselves: Measurebators spend endless hours debating and posting samples from each camera, pointing out the tiniest minutae of style differences so that other people can choose cameras that handle particular situations (specular highlight clipping being the most important to me).

I've been doing artistic and semi-professional photography for a decade and a half, starting by manually developing film from nikon bodies in the 90s, and then moving in to digital photojournalism early this decade. I know a lot of photographers, I am a photographer, and I will absolutely assert that "(digital cameras) compared to film are simply outmatched" is blatantly and patently false. The simple fact is that modern digital sensors capture a technically much larger envelope of information across the spectrum of imaging: More resolution, Larger colorspace, Higer signal to less noise with equivalent, or even Higher Dynamic Range, etc (not even touching the far cleaner workflow, the added advantage of immediate image review for stylistic learning, etc). The simple, linear exposure model of modern digital cameras is far superior for capturing the information actually in front of the cameras lens. With Digital, the color balance and dynamic range of your photograph does not depend on the precise temperature of the film manufacture, and/or developing, something you probably had no control over). Name a SINGLE axis on which Velvia or Kodachrome produce more information about the exposure than your modern full-frame digital sensor. Go on, try, I dare you.

Shooting Film is like shooting JPG, you lock in all of your style information and throw out a huge portion of your exposure data the moment the shutter closes. Maybe if your a luddite who lacks basic post-processing skills, you won't understand that this is all information you can trivially throw out later with the right curve, color balance and exposure profiles to produce the _exact_ same image you would have gotten with film, but at least you have the option to 'undo', or take the image in a completely different direction after the fact if necessary.

Film will remain a luxury niche product for well heeled aesthetic snobs stuck with their highschool dark-room days' photography expertise whom are willing to pay a premium so that they don't have to bother learning how to get the same results from modern technology, but the rest of the world has left yall behind.

Re:Nikon F6 and FM10

[Anonymous Coward]

> only two advantages SSD has over spinning media at this time: Access speed and Durability

and NOISE !!

Re:Nikon F6 and FM10

[Brett Buck]

Where is Kodachrome?

      Discontinued

Re:Nikon F6 and FM10

[AK Marc]

No one is going to say digital cameras are lousy, but compared to film, they are simply outmatched.

SLR digital cameras exceed SLR 35 mm cameras. The film stuff you are talking about is an issue of filters. Yes, you say you have crappy filters in software. But whether on the film or in software, it's a perversion. It's like the electric guitar amp. A digital amp would perform better in all ways, but they don't use them. Why? Because they want the amp to fail, and it's the manner of failure in the amp that generates what they are looking for. No mistake about it, it is a failure, but it is a specific controlled failure that is desired. Film is the same. It's the attempt to capture what the photographer wants, not the accurate representation of what's there, that gives any value to film. Film stores a perverted copy, but perverted in a preferred way. That isn't accuracy, that's creative freedom.

When you want your stored bits exercising creative freedom, then you'd have a point. But we want an accurate representation. And for that, your argument is irrelevant.

Spinning media is going to be with us for a while, and I expect, like film, that eventually prices will go back up and this technology will be a specialty market targeted at high-end users and professionals.

Aside from price, there is no benefit to spinning media. Regardless of price, some people like "old" audio and video tech. The old audio and video tech is "worse" in every measurable way (including price), but fails in desired ways. Storing bits can't ever do that. So, if SSDs were cheaper than spinning disks and available in the same size with the current parameters, I assert that we'd have no spinning disks (unless some tiny market happened to remain for places where mechanical disks were superior for environmental reasons, like being easier to shield against an EMP for military use or other such very specific and tiny markets). For general use consumer laptops, desktops and servers, we'd be 100% SSD.

Re:Nikon F6 and FM10

[maxume]

Anand paints is a little bit differently, he implies that (good) ssds should absolutely last 5 years, not just make it there:

http://www.anandtech.com/cpuchipsets/intel/showdoc.aspx?i=3403&p=4

And that's writing 20 GB per day to the drive.

So you might still be on that first SSD after all.

Re:Nikon F6 and FM10

[smchris]

Depends on your purpose. The latest edition of the Backyard Astronomer discusses how they used to like film and, for most uses, now heartily recommend larger-sensor digital for astrophotography. For a freakish demo, they compare a couple long-exposures of a valley scene at night so you can see the detail captured in the dark areas by digital compared to film.

Re:In 5 years

[VernonNemitz]

Spinning media won't die quite that quickly. Remember that flash drives still have a limit on the total number of writes; spinning drives don't. When they start making SSDs using a different tech, that has no limit on writes, then the death clock will really start ticking.

Re:In 5 years

[v1]

and if they find a way to get rid of the fans, that'll remove all moving parts from computers, which will be a good thing

Re:In 5 years

[Dragoniz3r]

Nor will they remember the days when they didn't have to replace their harddrive every 5 years because it ran out of rewrite cycles and the capacity started shrinking.

Re:In 5 years

[jeffmeden]

You can pry my spinning disk out of my cold, dead hands. There are a number of benefits still; cost and durability (aside from shock resistance) chief among them. Solid state disks will be taking a bigger role in storage, but spinning disks are a long way from being obsolete.

Tiered Storage

[Microlith]

The clock is certainly ticking, but it's got a long time to wind down. The largest barrier to the death of mechanical storage is the looming halt in NAND geometry shrinks, as processes get so small that it goes from being merely crap to wholly unreliable.

Seeing as how we've got 2TB in single disks now, and that capacity will likely continue to rise, I suspect we'll see capacity increases for SSDs slow for a while as new NVM tech comes online. Instead, prices will simply fall and you'll (hopefully) see some more consumer-oriented hybrid solutions where frequently accessed bits are stored in NAND and large, infrequently files will be out on your (hopefully RAID-6 protected) mechanical storage.

Re:Tiered Storage

[jmorris42]

> as processes get so small that it goes from being merely crap to wholly unreliable.

That hasn't stopped anyone from buying yet. Look at hard drives, the bigger they get the less reliable they are. Used to be if a drive survived the first couple of weeks it would probably end up on a shelf somewhere with a bunch of other tech that is useless but ya just can't bring yourself to trash because it still works.

> Seeing as how we've got 2TB in single disks now, and that capacity will likely continue to rise..

Or not. How long ago did the first 1TB drive appear? And we are still at 2 for the max? Methinks the hard drive industry has determined two very important things:

1. The demand for drives in the 1TB and larger size is fairly limited. But the relentless downward drive in prices is removing the profit from selling anything smaller.

2. SSD is increasing capacity per dollar purely on Moore's Law. SSD capacity increases leverage the general R&D investments in chip manufacturing processes. Creating a new generation of super hard drives would require the couple of surviving disc makers to expend Sagans of cash.

So we have a self fulfilling prophecy coming to pass where the drive makers have wound down R&D and are planning to reap profits while they still can. Existing drive tech can be squeezed to provide a couple more incremental capacity increases but eventually SSD will beat spinning platters in both capacity and price/GB.

This just in!

[Spazntwich]

Helicopters signal the end of automobiles, just as soon as their poor $$/mile traveled ratio reaches parity, but you can buy helicopters from Air Hog right now!

Solar panels signal the end of nuclear power AND the oil industry, just as soon as their poor $$/watt ratio reaches parity! But you can get a solar powered calculator RIGHT NOW!

Can I be a tech pundit yet?

Re:This just in!

[camperdave]

When was the last time anyone bought batteries for their calculator?

Re:This just in!

[Spazntwich]

Dammit man, I'm a slashdot troll, not a market researcher!

Re:This just in!

[TheLink]

Does buying UPSes for my computers count?

Re:This just in!

[hahn]

Helicopters signal the end of automobiles, just as soon as their poor $$/mile traveled ratio reaches parity, but you can buy helicopters from Air Hog right now!

Solar panels signal the end of nuclear power AND the oil industry, just as soon as their poor $$/watt ratio reaches parity! But you can get a solar powered calculator RIGHT NOW!

Can I be a tech pundit yet?

Yeah, and LCD's signal the end for CRT's...

Oh wait.

Re:This just in!

[blair1q]

No, you can not.

Those other things have inefficiencies built into them that cause their cost to remain relatively high.

There's no visible floor for the price of NVRAM. Currently the price of a device is high because the industry is recouping R&D. But as the learning curve progresses, the only way for them to keep the price of Flash die above the pennies-per-gigabyte level is to induce you to upgrade continually, the way CPU manufacturers do it. But even then, the pricing pressure is relentlessly downward. Intel and AMD have resorted to packaging 4 or 6 CPUs in one package just to get 20% of the price they used to get for one core that ran 30% as fast as each of the cores they're now shipping. Take a moment and research the cost of CF or SD memory cards over the past few years. I distinctly remember paying $200/GB and thinking it was a good deal (and it was, in 2002). Now you can get a 2 GB device with an MP3 player added on for $5.

I see no reason that SSD secondary storage won't follow the same slide into "need one? take one; got one? leave one" status.

So no, your tech punditry is the quality of vending-machine Borscht. No buzz for you!

Re:This just in!

[Lord Ender]

After upgrading one of my systems to a SSD, I went through and replaced all boot drives in all my systems with them. I can't stand using mechanical disks these days--waiting for things to load just derails my train of thought and damages my productivity.

These were the best speed upgrades per dollar I've ever spent. And thanks to you, I not only enjoy the performance boost, but also the feel of flying a solar-powered helicopter.

Re:This just in!

[Locutus]

No, not yet. You missed that the tech pundits are getting paid to throw this over the wall because of the relevance to Microsoft Windows. Remember how the first netbooks were all shipped with SSDs and GNU/Linux but when Microsoft came in and offered payment schemes for putting Windows XP on them, they all ended up with spinning disks? Windows just didn't fit very well/much on those small SSDs. Well, the SSDs are now almost ready for Windows so out come the dumbass stories to get people thinking about SSDs again. You'll soon be seeing them in devices with Windows and probably starting with netbooks. Remember, some pretty cool single and dual core ARM devices are heading to market this year.

The 'tech pundit' won't write about SSDs when all the fuss with SSDs is when they are the only storage means for GNU/Linux based systems. They don't get paid to do that.

You might not remember it but a number of years ago, an internal Microsoft email got sent out to the wrong address and it was learned that not only had Microsoft been hammering on a magazine editor to do a story on one of their projects but when that editor finally assigned a writer, Microsoft an atleast 12 employees assigned to _helping_ this writer put together his story. Some of those employees were given scripts to work from regarding conversations they were to have with this writer.

LoB

Re:This just in!

[Spazntwich]

Perhaps I failed at properly representing my point.

There is always some new technology signaling the end of current technology. There are always some hurdles that remain before this technology overtakes current tech because otherwise, well, it already would have.

Apparently, there is also always a market for tech writers who can sensationalize incremental price drops in consumer tech. HARD DRIVES WERE HERE TO STAY, FOREVER, UNTIL THESE NEW AMAZING PRICE DROPS HAVE FINALLY OPENED OUR EYES AND STARTED THIS AMAZING DEATH CLOCK FOR HARD DRIVES THAT NOBODY SAW COMING.

I'm not sure your specifics on the economics of consumer electronics are relevant.

I also disagree with your point on helicopters and solar cells. You could have said SSDs "have inefficiencies built into them that cause their cost to remain relatively high" a number of years ago and been correct about them. It would be naive to assume helicopters and solar cells won't undergo technological advances that result in significantly lower costs of construction.

Re:This just in!

[SunTzuWarmaster]

TI-89.

Just sayin'.

Re:This just in!

[Spazntwich]

Fuck you, man. I mean that in the nicest way possible.

Here I am, newly unemployed, bored out of my mind trying to have a good time mocking the folly of those I envy (someone getting paid enough to support a family by pretending he's John Madden announcing his team's gradual march down the supply curve) and you have to slap me in the face with a cynical expose of what turns out to be a well-considered and sinister plot to maintain the establishment's stranglehold over world markets of both products and ideas.

I'm gonna go play with my roommate's cat. And I hate cats. Microsoft probably operates entire underground breeding stations for stray cats. Every adopted stray and the ensuing "beer knocked over on laptop" is another license fee in their pocket.

Re:This just in!

[PCM2]

Pfff. HP 50g. Just sayin'.

Re:This just in!

[mtrachtenberg]

It seems reasonable to me that active data and programs (items used day to day) will be stored on SSD, and perhaps spinning hard drives will remain parts of systems with their role being to store all the programs and data that we want available, but only use infrequently or never. Perhaps the OS will only spin the hard drive up once on system startup (to build a directory on SSD), and then again if anything on the drive is actually needed.

Who really needs SSDs for Porn?

[Bananatree3]

SSDs offer speed. Spinning Disk HDDs offer cheap space. Hybrid disks offer a nice compromise until SSDs overtake spinning disks in storage/price.

I mean really, who needs an expensive big SSD for your porn collection? Unless you have 12 monitors running porn simulcasting...SSD speeds are really only needed for heavily accessed files. HDDs offer cheap storage for those not-so-often used files. The solution is relatively inexpensive, and here today

Re:Who really needs SSDs for Porn?

[masmullin]

.SSD speeds are really only needed for heavily accessed files.

Exactly!

Re:Who really needs SSDs for Porn?

[compro01]

That depends on the power of said oven and how long you feel like cooking for. A typical microwave oven can heat stuff up to over 1000C. Not enough to melt the platters, but most likely above their Currie temperature and would at the very least trash the drive electronics.

Re:Who really needs SSDs for Porn?

[hotkey]

I mean really, who needs an expensive big SSD for your porn collection? Unless you have 12 monitors running porn simulcasting...

Who are you and how did you get into my home?

Re:Who really needs SSDs for Porn?

[pebs]

Who (as an end user) needs any kind of storage medium for porn. Porn is in the "cloud" these days.

Re:Who really needs SSDs for Porn?

[rrohbeck]

I mean really, who needs an expensive big SSD for your porn collection?

Imagine how quickly you could access any random scene in your collection.

What about random writes?

[jtownatpunk.net]

Wake me up when cheap SSDs don't choke on random writes and their performance doesn't degrade significantly over time.

Re:What about random writes?

[MrNemesis]

If by "cheap" you mean the craptastic OCZ Core series, and the other SSDs of the same gen that used the god-awful JMicron controller, the "cheap" end of the SSD market is full of high-performance drives that don't choke on random I/O. As soon as Intel and Indilinx came out with controllers that were worth the sand they were printed on everyone started doing it, and we've now got a market where the performance delta between "cheap" and "prosumer" SSDs is much, much smaller.

The vast bulk of the cost of the SSD is the flash, which is why if you can find someone still selling Core series SSDs they're only marginally cheaper than an SSD that's actually usable.

Re:What about random writes?

[amorsen]

Here's your wake-up call: One of the two cheap drives mentioned is from Intel.

Re:What about random writes?

[jtownatpunk.net]

Sure, but the rated write speed on the "V" series is 35MB/s compared to 70MB/s on the "M" series and 170MB/s on the "E" series. It's not like you're getting last year's $400 drive for $125.

It's getting better and I'll probably give SSD another look in a year or so but there's still significant room for improvement when balancing price and performance.

Interesting assumptions

[jawtheshark]

The article seems to assume that a typcial laptop user needs a 120Gig harddisk. I don't think that's true. I can most certainly live with a 20Gig to 40Gig harddisk in a laptop. As a matter of fact, my current laptop (3 year old AMD Turion with "120Gig" HD) has two parts: about 16Gig fro WinXP MCE and the remaining 100Gig for Ubuntu. The 16Gig has all the productivity apps I need + 1 game (Portal), which still leaves me 2Gig free for data. If I didn't have the game, I'd have ~8Gig free for data. For typcial data like word processing documents and the like that is more than enough. It is perfectly usable for day to day tasks. (The Ubuntu part is my playground, but it could live just as wel on a 16Gig partition)

If you enter digital pictures into the landscape, it does change a bit. Still, that's still a lot of pictures. Besides, you don't want all your pictures on the move. They're much safer at home on server and/or NAS.

Music you say? We're talking about "needing"... You don't "need" music on your laptop, unless that's your profession, but that doesn't make you a typcial user.

While I don't think I'm going to shell out 100€ for a 32Gig SDD, because I'm a cheap bastard and what I have works, I could most certainly live with a 32Gig disk in my laptop.

Re:Interesting assumptions

[Yvan256]

Typical users keep music on their computer.

Re:Interesting assumptions

[MrNemesis]

IME the average laptop doesn't have WinCE, Ubuntu or games either. I don't expect your typical "home" laptop to start coming with an SSD for another year or two, but there's plenty of businesses that are drooling over the prospect of a) uber-fast laptop drives to make the loading of all that security bloatware bearable and b) laptop drives that are small and don't get stuffed full of files that should have been on the file server in the first place.

My "workstation" laptop has a 500GB drive, and I wouldn't give that up for the world - I need the space for VMs. But my knockabout laptop is a toshiba T110 with an aftermarket 60GB OCZ drive in it. I don't have to worry about it being jiggled, there's a noticeable improvement in battery life and, obviously, I/O responsiveness is much improved (which helps make the laptop feel faster than its wimpy single-core proc would have you believe). Extra space for media or similar is provisioned more easily in the way of USB keys or SD cards.

Moral of the story: IMHO people will go with a best-of-breed solution. Hard discs for people who need the storage, SSDs for people who don't. I think the article is right to assume that the majority of the laptop market can live inside 120GB.

Re:Interesting assumptions

[jawtheshark]

Well, people should at least back up their pictures on a NAS and/or USB/eSATA disks. If you don't have any you're playing with fire. Loss or theft of laptop means loss of all your precious data.

Also, if I get to repair a computer (laptop or desktop) the first thing I usually get asked to do is to save their precious digital pictures of their kids/vacation/whatever... because they have no backup at all.

My motto is: carry around what you need, keep on the NAS what you want to store and backup everything on external disks. (Which I don't do is have offsite backups, but technically, that would be the wise thing to do)

Best of all worlds and it won't set you back much financially these days...

Re:Interesting assumptions

[jawtheshark]

True... How much space does it use up? I once ripped my whole CD collection to MP3. Total was 9GB, half of which I never listen to. So your point is?

Re:Interesting assumptions

[jawtheshark]

WinCE?!? You must be confusing what I said: I have Windows XP MCE, which has nothing to do with WinCE.

I need the space for VMs

So do I on my work laptop... However, I was talking typical users, not computer enthousiasts and/or professionals.

I think the article is right to assume that the majority of the laptop market can live inside 120GB.

Of course it is, because it's major overkill. Even most Netbooks come with 160Gig HDs these days. I don't even think you can buy disks much smaller than that these days

Re:Interesting assumptions

[MobileTatsu-NJG]

Music you say? We're talking about "needing"... You don't "need" music on your laptop, unless that's your profession, but that doesn't make you a typcial user.

Fail.

Re:Interesting assumptions

[MrNemesis]

WinCE? Oh dear, brainfart :o

My point was that IMHO the typical user doesn't have VMs, games, Linux installs, blah - I thought from your post you were saying the assumptions were wrong and that the average user *would* need >120GB, rather than the other way around. Apols for the misunderstanding.

Re:Interesting assumptions

[kidgenius]

Total was 9GB, half of which I never listen to. So your point is?

My MP3 collection is 30GB and it pales in comparison to many others

Re:Interesting assumptions

[natehoy]

I feel the same way. My wife's ASUS eeePC Netbook came with a 350GB 5400RPM hard drive. It works OK, and of course an SSD would have jacked the price way up from the $300 price we got it for, but I'd still have preferred even a 16GB SSD drive on the thing. She'll never store any serious data on it.

It does have an SD expansion slot - I may at some point pick up a 16GB SD chip for it and see if I can boot Mint off the SD chip instead, then see if I can power down the hard drive in the BIOS as an experiment.

Re:Interesting assumptions

[c0d3g33k]

You must not be a developer. Even casual development (say you want to manage a few packages for your favorite Linux distro, or track some of your favorite projects via git) can eat up the space pretty easily. 'git clone' enough repositories or set up enough build environments or virtual machines and 100 Gig is gone quickly. Or maybe you are freelancing and need to keep the data for multiple clients on your laptop for site visits. I can't imagine what kind of space a busy graphics artist might need. Oh,wait. Maybe you have a dual-boot machine. 60 gig per OS if you split it evenly? Not enough.

This may not describe the 'very casual user' (aka my dog, because my mother-in-law is already too hardcore for your measly 120 gigs), but you don't need to be very hardcore nowadays to fill up 120 gigs. If that doesn't work for you, then just imagine a teenager with a serious interest in gaming.

Re:Interesting assumptions

[Art3x]

Yes, my work laptop, with Windows XP, is using just 12 GB. (But I don't have anything on it but Office and a few web browsers. But I'm a web programmer and do most of my work SSH'd into a Linux web server.) My home computer with Ubuntu uses, like, 4GB. I don't store pictures or music or movies on them, so I guess that's the difference.

Re:Interesting assumptions

[Mr EdgEy]

We're talking "average joe" here, I'm sure you know of people that use Limewire and download single 128kbps songs, they struggle to fill a 4GB portable.

Re:Interesting assumptions

[socrplayr813]

Definitely agree here. The vast majority of 'average Joe' computers that I come across have a 160-500GB hard drive, of which they've generally used less than 20-25GB, including all of their apps, music, etc. Unless they have an absolutely massive music collection or like to play a whole bunch of high-end games, a 40-80GB SSD is PLENTY.

Even a more tech-oriented person like myself doesn't need a ton of space. I keep all of my files on a server at home, which I access from my other computers when I want something. If I'm going on a trip, I can put the important stuff on my laptop, with all of my music and whatever videos/games I feel like I want for entertainment. That easily fits within my arbitrary 80GB and it would not take much effort to fit into 40GB. How much information can a person really need, especially for short trips? Are people really using 2TB of files on their laptop when they leave the house? (Short of possible work-related stuff)

Re:Interesting assumptions

[jafiwam]

Ignore me while I quietly stroke my 800 gb jukebox of MP3 music. [whistles innocently]

And I am not the biggest hoarder I know, it's about third, maybe fourth in file size.

Music aside, my ideal laptop hard drive size is 100 to 200 gb. Enough for decent entertainment and with a boot device, to store all the re-install files needed to get back up and running.

Re:Interesting assumptions

[t0p]

But do you really *need* to carry all 30GB of music around with you on your laptop? My music collection adds up to a pitiful 10GB or so, and that represents quite a few songs. I couldn't listen to all of them in one working day even if I wanted to. I've got one of those EeePC 701s, the one with the 4GB SSD. I'll put some music on it when I'm going out and believe I'll want to listen to some sounds. Meanwhile my entire collection is on a hard drive at home. Back in my youth, I had a Walkman. I also had up to 50 cassettes of music; and I *never* felt the need to take it all out with me. I'd slip a cassette or two in my pocket, and the rest stayed in my bedroom. Why would I feel the need to carry my entire collection all the time now?

Re:Interesting assumptions

[Idiomatick]

I don't keep music on my laptop either. Have 70,000songs on the server though... Spinning disks on that all the way. But I could see using an ssd in my laptop.

Re:Interesting assumptions

[nine-times]

Yeah, but how much music? I know some people who have several hundred gigabytes of music, which I find unimaginable since I only have around 12 GB. (It's not like I'm new to this stuff, but music is expensive and I don't pirate.) I can get by on a 40 GB hard drive in my laptop.

So definitely part of the question is, what's "typical"?

On the other hand, I'm just talking about my laptop; my home server is a different story. There I have >4TB of storage hooked up. So maybe there's a place for both for the time being.

Re:Interesting assumptions

[Beezlebub33]

Dude, I've got a 16G World of Warcraft directory! And a 5G music collection. And a couple of movies at several gigs for when I'm traveling (can't use DVD on plane, it dies too quickly). And a couple of gigs of pictures and videos from the phone. I've got at least 8G of development and testing work stuff (VS, and Eclipse, plus Tomcat, several database programs, log files, build and test directories, etc). I've got all the reports, documents, and presentations I've given for the past couple of years, in multiple versions (2 G). All told I've got 70G of 'stuff' that I don't really want to go delete, because it's just so convenient.

So, it depends on the user. For the average slashdot user, I'd say 30G is too small currently. 100 G feels about right, so you can have the stuff you need, and still defrag.

Re:Interesting assumptions

[Phred T. Magnificent]

That depends entirely on your usage pattern. It sounds like your laptop use fits the "carry on bag" pattern -- you throw a few files in the "carry on bag" laptop, go hop on an airplane, and sync up with the "mothership" later. (That's pretty much exactly the market that netbooks and the MacBook Air target.)

There are other patterns out there. Mine would be more the "Winnebago" usage pattern -- I certainly could keep some of my stuff on a server elsewhere, but I'd rather have everything right there with me and live on my laptop full time. (Sadly, I only have a 500 GB drive at the moment, but that may change depending on what happens with prices.) This also influences choices like the 17 inch screen and full keyboard.

I'm sure you'll find opinions to the contrary, but I'm going to insist that neither way is "wrong". Do what works, but don't expect the same thing to work for everyone.

Re:Interesting assumptions

[stewbacca]

Somebody whoosh me because I have no idea why this was modded +5 insightful.

Re:Interesting assumptions

[MobileTatsu-NJG]

Somebody whoosh me because I have no idea why this was modded +5 insightful.

"You don't really need a car, a telephone, or a refrigerator." (Score: 3, Insightful)

Do you get the point now?

Re:Interesting assumptions

[jawtheshark]

No problems :-)

Re:Interesting assumptions

[Yvan256]

Because for most people their laptop is their only computer and also their only storage space.

Re:Interesting assumptions

[linzeal]

I have a small music studio at home so maybe I am not the norm but I have over 100 GB of personal music in addition to 200 GB of music recordings/randomness I use for making music.

Do you rip in MP3 or FLAC is the question. My FLAC collection of ripped music is easily 70% of the total.

Re:Interesting assumptions

[stewbacca]

Still don't get it. I guess you are saying that music is no less needed than the other daily things in our life? I get that, but it doesn't have to require a huge hard drive on a lap top to do so.

Re:Interesting assumptions

[MobileTatsu-NJG]

No, that's not exactly what I'm saying. What you 'need' depends on what you bought a laptop for. When you start trimming popular things that lots of people do, like listen to music, and don't back that up with the benefits you'd get from doing that, you're not making the point that the solid state drives will take over.

If that's not enough, consider what a difference numbers make. One suggestion that they could easily outgrow the storage on their laptop and the purchase is toast. When the numbers go up, the masses are having more fun.

Re:Interesting assumptions

[owlstead]

Where are the laptops that offer a (1.8" or smaller?) SSD and a big slow inexpensive rotating platter, that's what I want to know (and where can I get one?).

And if anyone wants to produce this idea I have for a generic drive cradle that plugs into that slimline SATA DVD drive, I've got some initial drawings for free :)

Re:Interesting assumptions

[nine-times]

Do you rip in MP3 or FLAC is the question.

Well I've known people who have >100GB of music ripped at rates between 128kbps and 256kbps. They just had that much music. I just don't even want that much music on my laptop even if it were free. After a while, I feel like it just becomes too much to sort through and maintain.

But different strokes for different folks, right? Still, I don't know what's "typical".

Re:Interesting assumptions

[i_liek_turtles]

Your average user's only computer is probably a laptop. They're not going to mess with fileservers or externals.

Re:Interesting assumptions

[moonbender]

But do you really *need* to carry all 30GB of music around with you on your laptop?

No. But once you have 30 GB of music, you either keep all of it on your laptop or you need to decide which 2, 5, 10 GB of those 30 GB you do want. Which can be annoying. It's easier to just have it all available all the time. The same issue exists with portable media players. Note that I'm quite happy with my flash-based player, but I understand how people don't want to bother with managing a second "current" music archive. None of this is about need, incidently, the word is just meaningless in this context.

Re:Interesting assumptions

[WuphonsReach]

For our work, 120GB is about the minimum size for laptops. 40-60GB would be way too small for our tastes.

Personally, I need an affordable 500GB SSD before I can switch off of rotational media. Although I might try it when the 250GB SSDs finally drop below $300.

Re:Interesting assumptions

[dillon82]

I and many people I know stopped storing mp3s and now use only streaming audio sites a la grooveshark. Also long time lurker's first post on /.

not yet

[wizardforce]

The article seems to be making the point that if the average price becomes reasonable ie. 100 $$ then somehow people will start replacing their existing 500GB drives with 30-50 GB SSDs which is ridiculous. SSDs still need to be able to compete with hard drives at the $/GB level if they are to replace hard drives. Now that isn't to say that SSDs wouldn't have a niche like netbooks and SSD/hard drive hybrid setups but I seriously doubt we'll see SSDs take over the market in a few years.

Re:not yet

[maxume]

If people can buy an SSD with 'enough' storage for $150, lots of them are going to do it. That group of people is only going to grow.

(Personally, if there was a 1 TB SSD available for $2-300, it wouldn't matter very much to me what size HD the same dollars could purchase; I would think awful hard about a 500 GB SSD at the same prices.)

Reports of HDDs' demise greatly exaggerated

[kyz]

As TFA says, For $125 you get a 40GB SSD. Today on Newegg I can pay 110 for a 1500GB hard disk drive - that's about 40 times more storage, for LESS!

Unless SSD suddenly becomes 40 times cheaper, it's unlikely to wipe out regular HDDs. And it has to cope with the fact HDDs get better every year too.

There has always been a sliding scale in computing with "faster, less storage" on one end and "slower, more storage" on the other.

Cache RAM -- RAM -- Flash RAM -- SSD -- HDDs -- tape.

As time goes on, everything gets faster and everything grows in storage capacity - but they all stay the same relative to each other on the list. Anybody who tells you otherwise is selling you something.

Re:Reports of HDDs' demise greatly exaggerated

[courteaudotbiz]

And not to mention that 40GB is barely enough to have Vista or 7 breathing, once Office installed. Don't ever think about installing Windows 7 SP1 on such a tight space.

Oh, and don't you have those MP3s, Videos, Documents et al of yours?

Re:Reports of HDDs' demise greatly exaggerated

[Renegrade]

Totally agreed there, but you forgot "Registers" or "Register File" at the top of that speed progression.

Flash is going to run into the quirky nature of low-nanometer electronics before long, too.

Re:Reports of HDDs' demise greatly exaggerated

[ivan256]

How much for a hard drive that's as fast as that $125 SSD?

The 1TB Seagate hard drive that I recently tested gets random 4k read rates in the ~1MB/second range. My 80GB Intel X25-M gets ~38MB/second.

That's about 40 times more performance for THE SAME PRICE!

Storage capacity is irrelevant in many situations.

A 40GB SSD is more than sufficient for your average manager/executive. They'd almost certainly prefer opening Outlook and Power Point in a tenth of the time it used to take to having an extra thousand gigabytes of unused space on their laptop.

The 80 GB drive I have in my system was the best upgrade I ever bought. Kernel compiles are crazy fast, and all of the media I need can be streamed off the network (sharing a single one of those 1.5TB drives with a dozen or so other people).

Re:Reports of HDDs' demise greatly exaggerated

[Lunix Nutcase]

And not to mention that 40GB is barely enough to have Vista or 7 breathing, once Office installed.

And yet I have both Windows 7 and Office installed and it doesn't even take up 10 gigs of space.

Re:Reports of HDDs' demise greatly exaggerated

[zero0ne]

Wrong.

A full Win 7 Ultimate install with Office 2010 + Visual Studio 2010 + Project & Visio 2010 sits at around 25GB.

you still have 15GB left. Take off VS2010 and you are sitting around 20-25GB free.

Re:Reports of HDDs' demise greatly exaggerated

[gad_zuki!]

>As TFA says, For $125 you get a 40GB SSD. Today on Newegg I can pay 110 for a 1500GB hard disk drive - that's about 40 times more storage, for LESS!

Capacity is easy to get. I have a 500gig drive that I store stuff on and a 30gig SSD to run stuff on. I dont care about capacity, its a solved problem. I care about performance. My biggest bottleneck is my hard drive, that's pretty typical of most users. Its incredible what my SSD has done to my performance on a daily basis. Heck, once the price drops a bit on 60gb or 120gb SSD, I'm buying it and putting the 30gb in my laptop.

My Win7 boots up in just a few seconds, game maps load very quickly, apps load very quickly, latency is pretty much gone, etc. Its such an upgrade, I'm kicking myself for waiting so long. I'm not even one of those guys who is always upgrading. I have an old 1.8ghz C2D OC'd to 2.2ghz and an aging board to match.

I suspect once people get a taste for SSDs they'll consider spinning disks to be a pretty big downgrade. I never want to go back the same way I dont want to go back to single cores or 10mbps networks.

Re:Reports of HDDs' demise greatly exaggerated

[gad_zuki!]

>And not to mention that 40GB is barely enough to have Vista or 7 breathing, once Office installed. Don't ever think about installing Windows 7 SP1 on such a tight space.

Err, I have a 30gg SSD running Win7, Office 2007, Mass Effect 2, Battlefield 2, Bad Company 2, Red Alert 4, and about a dozen apps and random data (photos, documents) etc. I'm at 10gigs free now.

Re:Reports of HDDs' demise greatly exaggerated

[MrNemesis]

How much for a hard drive that's as fast as that $125 SSD?

It was probably a rhetorical question, but I'll answer anyway: lots. Lots and lots and lots.

We've in the middle of replacing 48U's worth of short-stroked fibre channel discs with 4U's worth of solid state drives. Capacity was never much of an issue with these databases (they only total about 800GB) but to get the performance with an IBM pSeries box cost stupid money - I don't know the exact figure but it was somewhere in the region of 50k a year just for maintenance.

Even if you're not using hardware like a pSeries, a SAN or disc array capable of sustaining >20,000 IOPS is still going to cost you silly money, take up alot more space and eat at least five times more power.

Re:Reports of HDDs' demise greatly exaggerated

[fulldecent]

Please watch this video for a demonstration of the *40 times more performance* available with SSD:

http://www.youtube.com/results?uploaded=m&search_query=ssd+hdd&search_type=videos&suggested_categories=28&uni=3

Re:Reports of HDDs' demise greatly exaggerated

[Rockoon]

OK, I have 1.5 TB's of storage space. Why am I going to buy another? Thats right.. I'm not..

I am guessing that your porn collection grows so fast that you are always shopping around for another terabyte or two? Thats not how the real world operates.

In the real world, we don't try to find someone retailing laptops with terabytes of disk space. In the real world, that SSD looks damn attractive in that stock laptop. Fast? yeah.. very fast.. Big enough? yeah.. big enough.. Wont turn into a worthless platter the moment I drop the laptop? yeah.. the G forces these things withstand is VERY impressive.. The laptop costs about the same as that other laptop with that mechanic drive? yeah.. prices are similar.

Welcome to the real world, rapid porn collector.

Re:Reports of HDDs' demise greatly exaggerated

[EvanED]

How much for a hard drive that's as fast as that $125 SSD? ... Storage capacity is irrelevant in many situations.

I don't think that the parent would dispute that SSDs have a lot of benefits over a magnetic HDD, including speed. But like capacity, speed is also irrelevant in many situations.

However, how fast of a hard drive do I need to watch a movie or listen to music? Less dramatically, how 'bout playing a video game? (HDDs would speed up loading times, but likely not the rest of the game.)

There's a lot of stuff that you don't need speed for, and space is more important. If someone has 500 GB of data, it doesn't matter how fast SSDs are if 500 GB of SSDs isn't affordable -- they'll have to decide either to lose data (unlikely) or have a magnetic HDDs.

What does all this mean? I think a lot of computers (at least many desktops) in a couple years will have both an SSD and magnetic HDD to play to their respective strengths. But it'll be a few years (I'd guess at least 5) before most desktops have only SSDs.

Re:Reports of HDDs' demise greatly exaggerated

[radish]

Where are floppy disks on your list? Or punched cards? Or paper tape? Or core memory? Or WORM optical drives? Or Zip Disks?

My point is that storage mediums do die when something better comes along. Spinning HDDs will go away, guaranteed. The question is when, and will it be SSDs (as we know them now) or something else which replace them?

Re:Reports of HDDs' demise greatly exaggerated

[PCM2]

Please watch this video for a demonstration of the *40 times more performance* available with SSD:

First, your link doesn't work. Second, I've recently replaced my boot drive with an Intel X25-M SSD. It's peppy enough, but in no way does it approach anything close to a 40x speed increase versus my old drive. That's pure hype. It boots pretty fast, which is useful when you have to reboot after Windows updates. But for day-to-day operations, your average computer user is barely going to notice the difference between an SSD and a traditional drive. Seriously. The speed increase feels about as dramatic as when you first ditched your old computer with IDE drives for a SATA-2 system. It's nice, but that's all.

Re:Reports of HDDs' demise greatly exaggerated

[Grishnakh]

Which drive is better completely depends on your workload. For storing giant media files which are always read sequentially, HDDs are obviously the better choice. A few milliseconds of access time isn't going to be perceptible when reading a single 4GB file, which only needs to be read at less than 1MB/sec to provide full frame-rate.

However, a database full of tiny records will perform much better with a SSD.

Re:Reports of HDDs' demise greatly exaggerated

[Grishnakh]

And it has to cope with the fact HDDs get better every year too.

Actually, no, they don't. The access times on HDDs haven't improved at all in years, and the sequential rates haven't improved much either. These are governed by physical characteristics: how fast the head can move to any point on the platter, and how fast the platter spins. Today's 2TB drives can't seek any faster than 10GB drives from a decade ago, and since they spin at the same speed, their higher areal density is the only thing that improves their sequential transfer rates.

The only thing that's improved drastically in HDDs is their capacity. That's only one factor in choosing a storage medium for a given workload.

Don't worry; I don't see HDDs going away any time soon for storing giant amounts of data, especially media (like HD video), as their sequential rates are very good and access time isn't important for these tasks. But for other things, I think we're going to see SSDs taking over.

Re:Reports of HDDs' demise greatly exaggerated

[petermgreen]

As time goes on, everything gets faster and everything grows in storage capacity - but they all stay the same relative to each other on the list. .
No they don't, different technologies advance at different rates, and sometimes a technology gets squeezed out completely because another techonology fills thier niche better.

For example removable magnetic disks are mostly gone ( a few people are still using floppies for legacy reasons ) replaced by USB flash sticks and to a lesser extent external hard drives and optical drives (CDRW drives had a chance to take over the removable media market for a while but they were hampered by the non-standard status of packet writing software and by the large physical size of the disks).

Magnetic tape hasn't been completely pushed out but afaict only the enterprise backup market really uses it these days due to the huge cost of the drives

While the concept of solid state drives is nothing new solid state drives that are competitive enough that many people are seriously considering replacing hard drives with them are. The probablly won't completely replace hard drives in the next five years but beyond that it's hard to tell.

Punched cards and paper tape are gone.

Re:Reports of HDDs' demise greatly exaggerated

[petermgreen]

One downside I see of the two drive setups is the user eduction requirement. I bet most lusers will think their computer is full when they fill up c: (i'm assuming c: is the HDD and a higher drive letter is the SSD).

Plus dual drives isn't really an option in laptops unless you go for the really big monster machines.

It's a good option for the power users but it's not a setup I expect to see much in off the shelf systems.

Re:Reports of HDDs' demise greatly exaggerated

[ivan256]

But like capacity, speed is also irrelevant in many situations.

However, how fast of a hard drive do I need to watch a movie or listen to music?

You don't even need a computer to do those things, so that's a silly argument.

When I say they're worth it, I assume you're doing actual work with your PC. If you're just going for music, games and movies, get yourself an iPod and an XBox.

Re:Reports of HDDs' demise greatly exaggerated

[ivan256]

My Intel X25-M on Newegg only cost $119.

Re:In 5 years

[delinear]

Depends on the purpose, really. As the basis for your OS, the number of writes might be an issue, but for general user data it's less so. I can see a trend developing in smaller hard drives to carry the heavy loads while data which doesn't require constant access is pushed onto increasingly larger SSD, and of course the move away from desktops to laptops and notebooks will drive this forward too.

Having said that, for home media servers it's not unusual to have several TB of linked hard drives, until SSD can even come close on both size and price, the humble hard drive should be safe for a while longer.

Re:In 5 years

[mattdm]

What makes you think that spinning drives don't have a limit on total number of writes?

32-40 GB isn't bad

[DragonWriter]

While most computers come with bigger disks (because the cost of making spinning disks makes the marginal cost bigger, and bigger numbers are always easier to sell), I've had 30-40 GB Linux setups on dual-boot machines where the primary was Windows, and never really had space problems. And lots of the things that eat up space on consumer machines (like video) are stuff that is better on a hard disk anyway. So I could easily see computers that aren't heavily used for video or similar applicaitons going to SSDs if 32-40 GB SSD are affordable, and computers with a 32-40 GB primary SSD as well as an HDD, where the HDD is mainly used for things where sequential transfer speed rather than random access time is key. The trick for the latter is getting a good configuration/UI setup that makes it "just work" for the most common applications without the user manually choosing locations (mapping locations appropriately, and maybe implementing MIME-type-based defaults for download locations), while giving power users precise control.

Re:32-40 GB isn't bad

[stonewallred]

30GB? Lol, my hot blonds with tattoos doing threesomes with brunettes with piercings and male chimps takes up more space than that, and that is a small niche classification in my collection....wait, did i just say that on /.?

Re:32-40 GB isn't bad

[rrohbeck]

Exactly. You can't make less than a platter worth of disk with some obsolete density, which means about 250GB per 3.5" platter and 120GB per 2.5" platter today. Some vendors may make 1/2 platter sized disks from drives with a failed head or leaving one head off, but with good yield the economics don't work out.
Oh and no matter what you do, the manufacturing cost of the lowest end hard drive will always be around $50. You can't make them any cheaper. A SSD OTOH, which is just a PCB, a few chips and some plastic, has a lot of room below that. Just look at USB sticks. They are SSDs too - just very crappy ones.

Re:32-40 GB isn't bad

[Kotten]

I agree

My laptop has a 30GB Vertex drive. I bought it a year ago because it was fast and cheaper/performance than velociraptor (which could not be put into the laptop anyway because of power consumption and heat). I run Ubuntu 9.10 64bit on it and using VirtualBox to run the other OS'es I need/want:

• Ubuntu 9.04 32bit
• Maemo SDK image (Ubuntu 8.04)
• Windows XP
• Haiku

Yes I am considering buying an Intel 80Gb SSD but OTOH I could ditch XP as I am not really using it.

PS. I stream music and film from the net or from my server

Re:In 5 years

[bluefoxlucid]

We know how to remove the fans. We have 18W 2GHz CPUs. As you roll these back in clock rate, power consumption drops. It takes n^2 power to run a CPU at a clockrate of 2 if it runs on n^1 power at a clockrate of 1; whereas if you have 2 cores, it takes 2n. When we drop power consumption by replacing spinning disks with 12V SSDs (not 3.3V fed, 12V at a third the amperage, less heat) and get low-power CPUs in there, the total dissipation will go away.

Re:In 5 years

[Anonymous Coward]

I think it'll be quite a few years before spindle drives are completely replaced.
$100 for a 20-40gb drive is still more than a 1000gig spindle drive.
I think spindle drives will be relegated to mass storage for media servers and other
such home devices. Big banks of personal storage using ZFS, Mirrored, RAID5 or whatever
ends up being used. I know the MPAA/RIAA wants us to stream everything (paying each time
of course) from the net but all those HD streams will block all the intertubes. With net
neutrality failing we'll see more and more bandwidth caps and rate limiting (unless you
pay for their premium internet service with faster streamed media for only $80 more).

Re:In 5 years

[oldspewey]

It's easy to get rid of all the fans in a computer. Just underclock it, and keep underclocking until you reach the desired thermal envelope.

Re:In 5 years

[Yvan256]

If you get rid of the fans, there won't be any funny/troll posts about Microsoft, Apple and Linux.

Re:In 5 years

[ArcherB]

Pity the lesson of Y2K went unheeded - where every COBOL programmer was paid whatever they asked to fix their code, but after should have all been taken out to a field and shot in the head.

Why shoot the programmers? Why not shoot the managers too ignorant to modernize their code base?

To get back on topic, I see spinning drives as the new backup or large file storage medium. You boot off your SSD and keep most of your files there, but anything you want a backup copy of or anything large enough to not need fast access, like movies, pictures, and music get stored on the HDD.

Re:In 5 years

[stonewallred]

Excuse my ignorance, but what is the security status of these things? Like running an erase HD command where you rewrite three or seven times for supposedly no data recovery. Are these similar, better, worse?

Ram Disks

[drumcat]

Oh yes... remember how RAM "disks" would soon rule once we got to a gig of ram, and all that extra ram was unneeded? There will be a time in the near future when you start seeing common augmentation (+1 for boobies being first).

Re:Ram Disks

[jtownatpunk.net]

I had a 2 meg RAM drive back in the 80s. It was freakin' awesome! The computer was up and running in seconds. Apps launched instantly with the touch of a button, like frogs in a dynamite pond. It had a battery backup that would keep the data alive through power outages of up to 6 hours.

Re:Ram Disks

[0123456]

I had a 2 meg RAM drive back in the 80s.

I had a 128MB RAM drive back in the early 90s; cost the company about $50,000 if I remember correctly. But boy did Windows 3.x start up fast.

Re:Ram Disks

[DragonWriter]

Oh yes... remember how RAM "disks" would soon rule once we got to a gig of ram, and all that extra ram was unneeded?

Um, no.

RAM disks -- being RAM -- aren't permanent storage, and were never held up as a replacement for any kind of permanent storage. They were always a work around for permanent storage being too slow for applications that had a demand for responsiveness and could accept the risk of data loss to acheive it, and even then were largely useful for applications that had been designed for older (relatively) constrained RAM situations, and therefore did stuff "on disk" not because it needed stored at that point, but to avoid keeping too much in RAM at once.

Re:In 5 years

[spazdor]

Yeah, if you really want to compare apples to apples, measure MTBF.

Oh, and let's not forget the SSD's far superior ability to decay gracefully.

Re:In 5 years

[EvanED]

SSDs could double in price/GB ratios every year for those 5 years and today's magnetic drives would still be about the same as what you'd got. 32-40 GB for $100ish now would mean 1-1.2TB for $100ish in 5 years with that doubling: but that's about what you can get magnetic drives for now. (Newegg has three or four 1 TB drives for $90, more than 1/4 less than the 40 GB SSD drive mentioned in this article.)

If hard drives double even once in that time, they'll still have a substantial price/size ratio lead.

Now, there's some point where drives become "big enough"; my disk space growth has slowed a bit from what it was a few years ago; I've basically only roughly doubled my use over the last 5 years. Combine that with the speed benefits of SSD and they'll definitely be much more mainstream in that time. 90% chance I'll have basically all my non-music, non-video, non-photo data stored on SSD in 5 years. 99% chance I'll have all my non-music, non-video, non-photo, non-large-video-game data stored on an SSD at that point.

But will I still have magnetic drives in my system in 5 years? 90% says yes to that too.

Re:In 5 years

[mrsteveman1]

I've never seen a consumer hard drive last even 3 years, some of them start clicking within 6 months, others spit out bad data after 12 months making them essentially useless even though they're still spinning (and therefor not "defective" for warranty purposes).

SSD Drives

[CFBMoo1]

The SSD drive on my netbook has been running great ever since I bought it. I'm really looking at buying a SSD drive and turning my newer SATA platter drive in to an external backup drive of sorts. Since I won't need to access it very regularly it'd make the perfect backup tape of sorts.

80 gigs for $224 ain't bad...
http://www.newegg.com/Product/Product.aspx?Item=N82E16820167023

Re:SSD Drives

[CompMD]

"80 gigs for $224 ain't bad..."

I beg to differ. That's more than I paid for my entire netbook, and I got twice the storage.

Re:In 5 years

[mrsteveman1]

What makes you think spinning drives are durable? They're already heavily reliant on ECC to reproduce digital data from an analog medium, you think they're going to continue to be reliable as the areal density increases? I don't.

Re:In 5 years

[LaughingCoder]

Interesting. Hard drives replace tape backup. SSDs replace hard drives.

Re:In 5 years

[lorenlal]

I have. I still have a hard drive from 1992, running on a 486/33, running an old Redhat distro..

Why? Because I can.

Re:In 5 years

[0123456]

I've never seen a consumer hard drive last even 3 years

Maybe you're doing something wrong in that case, because all but one of the five consumer drives in my Windows PC are over three years old and it's still working about as well as a Windows PC ever does.

And personally I've never bought a drive which failed in less than three years (for that matter I've only ever bought one drive which failed before I swapped it out because it had become too small).

Re:In 5 years

[c0d3g33k]

You must not get out much, since I have drives on my home file server that are close to 10 years old and still running fine. Or maybe you need to stop buying your drives from the clearance bin.

Extra tinkering still required

[jones_supa]

There's the reliability issue but then I find the babysitting annoying. Firmware updates, performance refresh utilities, partition alignment... With HDDs you didn't have to worry about any of this. I hope with future SSDs neither.

Re:Extra tinkering still required

[0123456]

Firmware updates, performance refresh utilities, partition alignment... With HDDs you didn't have to worry about any of this.

I take it you've never bought an HDD with 4k sectors?

Re:Extra tinkering still required

[Grishnakh]

If you're running Linux, this isn't a problem. It "just works".

Why does one have to win?

[wisnoskij]

While I see SSD becoming more common in the future, if HHDs continue to be significantly cheaper for large storage they will continue to play a roll.

While SSDs are getting cheaper and cheaper, computers are needing more and more storage space.

Right now I have 2TB installed on my desktop and if I had wanted SSDs it would of cost $7000~=((2,000/44)*$125).
Which is obviously not even close to affordable.

SSDs make a lot of sense for some things but are not likely to replace HHDs anytime soon.

Re:Why does one have to win?

[Rockoon]

While SSDs are getting cheaper and cheaper, computers are needing more and more storage space.

Not really.. we blew by the 130GB barrier (you should know what that was, or get off the lawn) 8 years ago, yet most people still only need 40GB or so.

Most people do not need lots of space. They have the OS, a few productivity apps, a web browser, a media player, their MP3 collection, and photos from their camera. Thats it. Easily fits in 40GB. Most people are not ripping DVD's, or using torrents, or packing 100 games onto their drive, a collecting massive amounts of porn.

All this talk about storage costs... if you only cared about storage costs, you'd be running tapes, not platters.

Re:Why does one have to win?

[wisnoskij]

I have no idea about the statistics about most people but everyone I know needs more then 40GB.

"or packing 100 games onto their drive"
Lately it seems all big name games are over 10GB per game, meaning you would not even need to be a avid gamer to use more then 40GB on just games installed on your computer.

Re:Why does one have to win?

[Rockoon]

Most people do not install games, at all.

Re:Why does one have to win?

[petermgreen]

All this talk about storage costs... if you only cared about storage costs, you'd be running tapes, not platters.
We don't ONLY care about cost per gigabyte. We care about getting the right balance of price, capacity, transfer rate, latency and IOPs* for the workload we plan to use.

modern tapes do indeed have the lowest cost per gigabyte for bulk storage but they are let down by insane seek times, high drive costs (which means the cost per gigabyte for smaller ammount of storage is high). Afaict they aren't bootable either (not that you would want to with the kind of seek times a tape drive offers). In other words they aren't a practical option for ordinary computer users.

SSDs meanwhile have the highest performance but thier cost per gigabyte is far higher than hard drives as is the cost of entry.

Hard drives are the option that offers a low cost of entry, more storage than most people will need even with the cheapest products and performance that is acceptable most of the time.

Are the people who only use their computers as glorified word processors and internet terminals (and the corporate buyers who buy for them) really going to pay extra for a SSD? I doubt it unless there is a REALLY big marketing campaign to convince them and maybe not even then
.

* average latency and IOPs are directly linked for a drive that can only do one operation at a time but not for devices that can do multiple operations at a time.

Re:Why does one have to win?

[Rockoon]

Are the people who only use their computers as glorified word processors and internet terminals (and the corporate buyers who buy for them) really going to pay extra for a SSD?

But its not extra cost... its just smaller. Thats the point.

Re:Why does one have to win?

[petermgreen]

But its not extra cost
Bullshit have you looked at the price of drives arround the 320-500GB mark (which seems to be what you mostly see in off the shelf PCs at the moment), it's less than $50

Plus I think 40GB is a bit on the small side even for basic users, how big is an install of windows 7, whatever the latest version of office is and whatever crap either the OEM or corporate IT want to load on the machine?

Re:Why does one have to win?

[Rockoon]

Bullshit have you looked at the price of drives arround the 320-500GB mark (which seems to be what you mostly see in off the shelf PCs at the moment), it's less than $50

Cheapest 320GB laptop drive on newegg is over $50, besides which you still missed the point. Normal people dont buy hard drives, they buy systems.

Let me repeat that for you. Normal people do not buy hard drives. Normal people buy systems.

OK one more time. Normal people do not buy hard drives. Normal people buy systems.

Systems are not configured and then a price decided upon. Thats not how it works. A price is decided upon and then the system is configured. This is because the market will bare a certain price, and thats what is targeted. The box with the SSD and the box with the HDD will have the same price range, but offer different features. The SSD box will offer performance, the HDD box will offer capacity.

If you dont believe that this is true, then explain why so many notebooks and netbooks are sold with SSD's in them, right now.

You are arguing for the geeks who will buy a hard drive and know how to install it. Geeks do not matter. Geeks are not the market. The geek market is irrelevant. You are not normal. You do not define the market. You are a unique and ugly basement dweller that does not define the market. Your opinion of what is and is not a good deal is irrelevant.

As far as normal people having office.. umm.. no.. normal people do not have office. Normal people have Word. They do not need a spread sheet or a database.. in fact.. they dont know what that shit is, and if they had them, wouldnt know how to use them.

Stop arguing like your grandmother thinks like you. She doesn't. Nor do your parents, or their friends, or their co-workers.

Re:Why does one have to win?

[petermgreen]

If you dont believe that this is true, then explain why so many notebooks and netbooks are sold with SSD's in them, right now.
Are they? Netbooks intially shipped with tiny slow SSDs but virtually every one on the market now has a hard drive. As for full sized laptops hardly any of them come with SSDs.

As far as normal people having office.. umm.. no.. normal people do not have office. Normal people have Word.
Maybe this varies with location but I don't think i've ever seen a standalone copy of word (I have seen "works suite" on occasion which had word but not the rest of office but only as an OEM bundle). People either have whatever shit was bundled with their machine, have office or have openoffice. Of course some of them may only use the word processor part of it...

We shall see how this pans out but I still think a vendor who ships a machine with a drive this small is setting themselves up for a world of hurt when customers see their brand new machine (with win7, office trial, antvirus trial and all the other stuff in a standard OEM image) is already over half full.

Re:In 5 years

[nxtw]

Just add more error correction coding.

Re:In 5 years

[Dan Ost]

Are hard drives really price comparable to tape?

Re:In 5 years

[MBGMorden]

You are definitely doing something wrong. I've had hard drives fail myself - heck I had a bout with an IBM Deskstar where it and it's replacements kept failing within 6 months. HOWEVER, the vast majority of them work much longer. I've got an old 80MB Connor hard drive from my original 486 that still worked last time I plugged it in (about 2 years ago), and I've got at least 3 drives in normal usage that are 8-9 years old.

Just get one...

[Bearhouse]

Seriously. I've got small ones on my main machines, (just the OSes and pgms - plenty big enough) plus a bigger one on the laptop.
Internal or external classic HDDs give plenty of cheap space. With SATA, even external drives are fast enough.
Forget about springing for the latest multi-core gonzohertz CPU; these things have make a real difference to everyday usability.
OK you only boot once per day, but application and big datasets load fast...laptops hibernate and reload fast too...nice.

Re:In 5 years

[0123456]

I have. I still have a hard drive from 1992, running on a 486/33, running an old Redhat distro.

Good point, in 2007 I was copying files off my 1995 laptop drive before I threw it out; though I did lose a few files due to bad sectors.

Re:In 5 years

[0100010001010011]

Tape drives are still alive and kicking. Maybe not in your Commodore 64, but it's a great way to back stuff up. Many businesses still use them. Tapes are cheap and are quite a bit more hardy than disks or DVDs.

Re:In 5 years

[metamechanical]

The only real issue there is that it couldn't really be intended for long-term, infrequently-accessed storage. A prominent failure mode of platter drives comes (surprisingly) from sitting idle, not being used.

Damned fast, worth the price

[ka9dgx]

I received a 128 Gb Kingston SSDnow as a gift from a friend, to put in my laptop. The laptop had a 320 Gb hard drive, so I've had to not lug 2 years of photos around, but it's well worth it because this this is damned fast. Things that had 10 second times now are sub-second. The thing boots Windows 7 in less than 10 seconds.

Capacity is nice, but once you get past 40Gb or so, you only need it to store images and things in bulk. It's like having the speed of a SAN in a laptop. SSD is an order of magnitude faster as far as the user experience goes, and if you can get one for less than $200, it's well worth doing, IMHO.

Once the end users see this in action, the price/Gb won't matter to them, because responsiveness is the name of the game.

Re:Damned fast, worth the price

[zeet]

Lens:Photographer::Pen:Author

Some of the most stunning photographs in history were taking with equipment that would make the measurbators cry. All the sharpness in the world cannot make a boring picture into an interesting one.

Re:In 5 years

[bsane]

(not 3.3V fed, 12V at a third the amperage, less heat)

P=IV

So I'm having a hard time reconciling how raising the voltage by 3x (roughly), and using a third less current changes power consumption at all. I'm pretty sure transmission distances and losses are pretty low inside a computer case.

I think so.

[SanityInAnarchy]

Hard drives are a terabyte for under $100. Tapes even close to a terabyte seem to require drives in the multiple thousands of dollars, and I can't find prices on the media. Even if you factor in some sort of hot-swap chassis and a server to run it on, I think hard drives win.

Re:I think so.

[Amouth]

LTO5 tapes beat spinning disks in the cost per gigabyte for backups - BUT they require a 6-7k drive to use and are far slower than disks.

while for the enterprise that burns through tapes for backup s a autoloader and LTO drive/tapes is the best solution

for the small company (2-3Tb of data backed up weekly or monthly) a rotation of large drives works fine and isn't nearly as expensive

Re:I think so.

[Forge]

Dell sells LTO-4 (800GB/1.6TB) for $50.45 with the purchase of a drive. Since the drives start at $3,249.00 You need to be using around 60 tapes before it matches the price/GB of a sub $100 1TB SATA drive.

More than two years ago the balance shifted. It is now cheaper to build massive storage servers with SATA RAID in-house and off-site and backup to both than to put a Tape Library in your office and rotate tapes off-site.

This is true even when you assume $0 for transporting tapes and free off-site storage.

Re:I think so.

[BlackSnake112]

Remember to back up twice (two drives) if going the hard drive route. I have seen too many people who backed their data up to a external drive that failed.

Re:I think so.

[SanityInAnarchy]

Dell sells LTO-4 (800GB/1.6TB)

In other words, 800 gig, so they assume 1.6 terabytes will fit once compressed. Hey, I can do the same thing for a hard drive, count that 1.5 TB as 3 instead. But let's stick to actual physical size...

1.5 TB hard drives run about $99 on Newegg, or about an extra 3 cents per gig, or about $2.35 per drive. Unless my math is off, that's well over a thousand tapes to pay for the drive.

In other words, this starts to pay for itself when you have to store over 800 terabytes (and probably closer to 1 PB) of data. At that size, yes, transportation and storage starts to matter, not to mention maintenance and basic operations -- maybe you're right, but at that point, I have no idea whether the human to swap tapes and physically move them (or the human in the remote tape library to handle them after sending data over the Internet) or to run that many extra servers (where do all those hard disks go? Do they spin down when not in use?)

Re:I think so.

[trajik2600]

Remember to back up twice (two drives) if going the hard drive route. I have seen too many people who backed their data up to a external drive that failed.

You've never seen a tape restore fail?

Re:I think so.

[BikeHelmet]

I got a 1.5TB WD Green drive for $99 CAD.

At the time that was about $90 USD. You're right - they're quite affordable, as long as you only need a couple dozen terabytes or less.

Re:I think so.

[bill_mcgonigle]

Hey, I can do the same thing for a hard drive, count that 1.5 TB as 3 instead.

This is not a fair assumption. With hard drives, you're probably going to be running ZFS with compression and block-level de-dedup turned on, so for many usage scenarios you should assume 6 or 8TB.

Re:I think so.

[mysidia]

Actually... this just goes back to test your backups, regularly, make sure you read back and verify every bit on the medium.

Even if you use tape, you need to have two backups at all times anyways -- otherwise, what do you do for backups if your source media blows up in the middle of a backup, or starts spewing corrupt data (during read operations), giving you both an unusable source, and an unusable backup destination?

Re:I think so.

[dbIII]

One or two due to actual media problems out of what must be around a thousand by now.
Of course there's been a few cases where it was never on the tape on the first place.
Of course the same two tapes deal is the way to go with anything really important.

Re:I think so.

[mjwx]

1.5 TB hard drives run about $99 on Newegg, or about an extra 3 cents per gig, or about $2.35 per drive. Unless my math is off, that's well over a thousand tapes to pay for the drive.

LTO-4 are 800 GB uncompressed, 1600 GB compressed, I get a compression ratio of about 1:1.2 - 1:1.4 So I fit between 1 and 1.2 TB on each tape, some data compresses up to 1:1.8. But that's not the advantage of tapes.

Hard drives are extremely fragile and not portable. Drop, mishandle, submerge, freeze or run over a harddrive and then recover data from it. With a tape you can destroy the casing, even cut the actual tape and you can still recover from it. Maintaining a large RAID array on-site doesn't help if your building burns down and maintaining one off site is quite a bit more expensive then a 24 LTO-4 tape loader.

You fail to take into account the cost of actually buying and maintaining a redundant RAID array. An EMC SAN will cost upwards of US10K, an IBM ServerRAID 8i PCI-X card will cost about US$800, even software raid requires a box with an OS to run it with warranty and more hours spent on maintenance. Even if you're mad enough to run a critical business system on consumer hardware a decent PC will set you back A$1500 for a 6 disk system, A$2000 for a 7 disk system as most consumer boards only have 6 SATA slots. Aside from the risk of running consumer disks (3yr warranty) compared to enterprise level tapes (lifetime warranty from IBM and Tandberg) there is little difference in the set up costs and a big difference in the cost of maintenance and life expectancy.

On-line backups are difficult when your data upload rate is limited, I back up 6 TB weekly plus daily differentials which are about 1 TB per week, one set of tapes goes off-site each month. unlimited 10 Mbit Fiber costs A$1500 a month in Australia, 8 LT0-4's cost A$560 and copy the full 6 TB in under 18 hours, that 10 Mbit link would take 52 days to copy that data at it's theoretical maximum transfer rate. Yes we wont have to transfer the entire lot each week if we use a differential or incremental backup but we backup 1 TB a week just in differentials so that's still 8 days. 75 MB/min max theoretical transfer rate for 10 Mbit fibre to the 3.6 GB/min real world I get out of a LTO-4 tape. Granted I get close to the max on the 2 Mbit fiber line we have (A$400 per month).

Tape is cheap, fast, reliable and high capacity. Other backup solutions are cheaper or faster or have higher capacity or more reliable but never manage to combine all these qualities like tape. Optimality I'd like to combine differnt backup solutions so they are redundant and I don't have a single point of failure, but I'd also like a pony.

Re:I think so.

[SanityInAnarchy]

LTO-4 are 800 GB uncompressed, 1600 GB compressed, I get a compression ratio of about 1:1.2 - 1:1.4 So I fit between 1 and 1.2 TB on each tape, some data compresses up to 1:1.8. But that's not the advantage of tapes.

Um... duh?

In fact, I would much sooner turn off the hardware compression (assuming that's what they're doing) and do it in software -- probably get a better ratio, or tune it based on how much CPU I feel like burning on it.

Hard drives are extremely fragile and not portable.

More fragile than a tape, maybe, but they're certainly portable.

With a tape you can destroy the casing, even cut the actual tape and you can still recover from it.

You can do the same with a hard drive, it just costs more.

Maintaining a large RAID array on-site doesn't help if your building burns down and maintaining one off site is quite a bit more expensive then a 24 LTO-4 tape loader.

The former is obvious, but I'm not sure about the latter -- hosting is cheap. How long would it take for the tape to pay for itself? And that's assuming it's cheaper to hire someone to physically move the tapes offsite.

You fail to take into account the cost of actually buying and maintaining a redundant RAID array.

Not really, I just assumed it would be done intelligently. In other words:

An EMC SAN will cost upwards of US10K, an IBM ServerRAID 8i PCI-X card will cost about US$800,

Overkill.

even software raid requires a box with an OS to run it

Linux is free. So is Solaris, for that matter. And a box can be had for less than $800 -- assuming you need a dedicated box that will do nothing else, which again seems like the dumb way to do it.

Even if you're mad enough to run a critical business system on consumer hardware a decent PC will set you back A$1500 for a 6 disk system, A$2000 for a 7 disk system as most consumer boards only have 6 SATA slots.

Are you factoring in the cost of the drives themselves? If so, I don't see how a drive and an extra controller (even PCI will do) is an extra $500.

Aside from the risk of running consumer disks (3yr warranty) compared to enterprise level tapes (lifetime warranty from IBM and Tandberg)

Hence the RAID -- and chances are, by the time any of them fail, you'll be able to replace them with something much larger for the same price. If you're running ZFS, it doesn't even have to be the same size.

Then again, why are we doing RAID? This is backup. Do checksumming and call it a day -- if a file is lost in the backup, we presumably have a live copy anyway.

there is little difference in the set up costs

Yes -- though it gets much cheaper if you pay someone else to do that. For example, Amazon S3.

On-line backups are difficult when your data upload rate is limited, I back up 6 TB weekly plus daily differentials which are about 1 TB per week, one set of tapes goes off-site each month.

Data upload rate, and in Australia -- ah. Now it does start to make sense -- so you're right, you'd need to build the offsite facility yourself, and you'd probably need to physically transport anyway. Given that, tapes start to make sense.

I am a bit skeptical of your daily diffs, though. Where does that amount of data churn come from?

Re:I think so.

[jridley]

I've seen more tapes fail than hard drives. Be sure to make two tape copies as well.

Re:I think so.

[Areyoukiddingme]

I am a bit skeptical of your daily diffs, though. Where does that amount of data churn come from?

And why hasn't he ever heard of rsync?

Re:I think so.

[mjwx]

In fact, I would much sooner turn off the hardware compression (assuming that's what they're doing) and do it in software

That I already do. The rate of compression depends on the data type, as I'm sure you already know.

More fragile than a tape, maybe, but they're certainly portable.

How, HDD's are heavier and more fragile then LTO 4 tapes, also more vulnerable to static. You simply have to put a LTO tape back in its case, with HDD's they have to be in an anti-static bag and impact resistant box (we are talking about your entire business if your office burns down, you need to take care). I've had bad experiences with HDD's, properly stored for years in anti-static bags then plugging them in and having them not work, even a failure rate of 1 in 20 is too high.

You can do the same with a hard drive, it just costs more

Exactly, smash tape with a hammer and just roll the tape into a new case, smash a HDD with a hammer and pay through the nose for recovery as long as the plates are intact.

hosting is cheap.

For 6 TB? Add to this the cost of a 100 Mbit fibre connection to move 1 TB a week (lets ignore the entire 6 TB full and concentrate on the data that changes). Next you contradict yourself a lot.

And that's assuming it's cheaper to hire someone to physically move the tapes offsite.

You assert that tapes require a enterprise level service to be properly stored (well actually it can be done for A$20 a week for a fireproof lock box).

Overkill.

Then you blatantly state that enterprise level hardware for your scenario is overkill.

Pick one standard, apply to both scenario's otherwise your using a strawman.

Linux is free. So is Solaris, for that matter.

My time however is not. Compared to my time a Windows license is almost free. The cost in your scenario comes in maintenance.

Are you factoring in the cost of the drives themselves? If so, I don't see how a drive and an extra controller (even PCI will do) is an extra $500.

A$2000 for a box, with controller and 1 disk. Add another A$150 per 1.5 TB disk. Unless your suggesting we should use the receptionists machine as a backup system. This is where your system becomes highly unreliable. Not enterprise ready at all.

Hence the RAID -- and chances are, by the time any of them fail, you'll be able to replace them with something much larger for the same price.

Here is the problem, you're treating your backups as a hot system, not a cold system. I've lost count of how many times someone has asked me for a file they deleted off the server months ago, do you intend to keep adding disks to this array?

Further more, your idea of scalability is bad, "by the time"? You're gambling on technology that hasn't been released or perhaps even invented yet. LTO-4 is already a system that is highly scalable and cheaply scalable.

if a file is lost in the backup, we presumably have a live copy anyway.

Presume nothing. Assumption is the mother of all fuck ups. Backup systems must be designed so that there is as little presumption as possible, you have the live copy, the last edited copy and the last four weekly copies sitting on tape, not to mention several historical copies moved off-site. To keep all this live is risky and costly.

I am a bit skeptical of your daily diffs, though.

A differential backup is all data that has changed since the last full backup, the archive bit does not get reset to after each diff so the size keeps growing until the next full backup.

So I think that backing up 1/6 of your full in 5 days is not unusual, actual data changed will be less then that as this diffs double up a bit. I work in a GIS business, large imagery (gigabytes per image), if you make one change to the file you back up the entire file.

Re:I think so.

[SanityInAnarchy]

HDD's are heavier and more fragile then LTO 4 tapes, also more vulnerable to static.

So not as portable, I know, but nonetheless portable. I have an external hard drive which sat in my backpack for several months, and I carried that backpack everywhere -- the drive is still good.

(we are talking about your entire business if your office burns down, you need to take care).

Yes, that's true, but it's also something you'd presumably be monitoring. The idea of a backup is to have at least two copies, so that if one copy fails, you can reconstruct from the other. If the "portable" hard drives fail, you buy a new one and copy the data from the original source -- the live system.

hosting is cheap.

For 6 TB?

Depends how you do it. With a colo, I imagine it would be -- just make sure it fits in the box.

You assert that tapes require a enterprise level service to be properly stored (well actually it can be done for A$20 a week for a fireproof lock box).

I suppose I'm making assumptions about the physical size of the tapes, and the need to have them organized.

I'm not assuming "enterprise-level" -- those are your words, not mine. I'm factoring the labor of actually physically carrying the tapes and storing them, even if it is one person carrying them to a lock box.

Then you blatantly state that enterprise level hardware for your scenario is overkill.

For the drives? Yes. You'd want to put them in some rented corner of an enterprise-level datacenter, but you don't necessarily need enterprise-level disks and controllers, even to move a terabyte (or six) a week.

Linux is free. So is Solaris, for that matter.

My time however is not. Compared to my time a Windows license is almost free.

Irrelevant, unless you're claiming it's going to necessarily take less of your time to do it with Windows.

A$2000 for a box, with controller and 1 disk. Add another A$150 per 1.5 TB disk. Unless your suggesting we should use the receptionists machine as a backup system. This is where your system becomes highly unreliable. Not enterprise ready at all.

"Enterprise-ready" is a bit of a buzzword. Google runs largely on consumer hardware. I'm not suggesting you literally take the receptionist's machine...

And even with the exchange rate, sorry, 1.5 TB is not A$150. Try more like $110. Not "enterprise-ready", you say? That's another reason to use something like Solaris and ZFS -- it doesn't have to be. Even RAID stands for "Redundant Array of Inexpensive Disks."

Here is the problem, you're treating your backups as a hot system, not a cold system.

I'm not sure what you mean by that.

I've lost count of how many times someone has asked me for a file they deleted off the server months ago

So use hard disks and something like... oh... last time I set anything like this up, I used BackupPC. At this point, though, even a straight copy would benefit from block-level dedup. Either way, the result is that you can store many versions -- daily backups for a week, weekly backups for a month, monthly backups for a year...

If they deleted it off the server months ago, there's probably a version of it somewhere. Maybe not the most recent version, but a version.

do you intend to keep adding disks to this array?

Possibly, but that's not required by the above. If it were (assuming your massive numbers of changes), there's always the possibility of taking a snapshot of the system to a disk or five (using something simple, like a tarball split across those disks) and unplug those disks for awhile.

Further more, your ide

Re:I think so.

[Forge]

I used my own experience working for a company where:

1. A hour of lost Data from certain machines may mean a million dollars in lost revenue.

2. We own the raw fiber between locations.

3. We also have live systems at the backup site.

Even so, this is what I suggested for our next generation of backup hardware. 40TB of usable backup space (after RAID and hot spares) for under $16,000 is nothing to sneeze at. Being able to double that for an extra $10,000 or much less if you get the hard drives somewhere else ($16,000 for 80TB if you combine this box with drives from another source)

Before you ask, I checked. The machine comes with the empty drive trays if you don't order drives.

Re:Tiered Storage - Software joining SSD+HDs?

[thms]

Tiered Storage - "on the fly" support for that is something I would love to see:

• All your files are on on big 2TB rotating-rust* HD
• The OS automatically uses a 10 or 20 GB SSD to cache the parts which are accessed most.
• Then of course the usual RAM cache on top of that.
• And CPU cache on top of that, thus - 4 tiers!

I'd prefer that software solution to a hardware solution since the OS knows so much more about which files it would make sense to cache and which aren't worth it. Also, you could overrule the prediction algorithms easily to cache the music you want to listen to or the database you are working on. I actually use /dev/shm (a Linux tmpfs in RAM) often to store quickly changing files.


* I know iron oxides aren't used anymore, but I still like the mental image :)

Not so fast.

[amn108]

Don't be fooled, people: http://en.wikipedia.org/wiki/Solid-state_drive#Disadvantages

Re:Not so fast.

[MobyDisk]

I'll edit the article later. For now...

- Wear leveling used on flash-based SSDs...
Oops! Wear leveling is done on HDDs too. And it isn't a disadvantage: it is a solution.

- More expensive, lower capacity
Don't need to address that, since that is the topic of the article...

- Asymmetric read vs. write performance
Oops! Platter drives have this problem too!

- Requires TRIM
Solved.

- Limited lifetimes
Funny, that's considered the major downside to platter drives. Anyway, this is the same as the first point.

- Performance of SSDs degrades with use.
Solved. See TRIM. Note that this is also a problem on platter drives.

- SATA-based SSDs generally exhibit much slower write speeds.
This one doesn't even make sense.

- DRAM-based SSDs
Aren't what we are talking about, so that's irrelevant.

Re:Not so fast.

[MrNemesis]

If I hadn't already spammed this thread with posts I'd have modded you up - other than price/GB, most of the problems in the wiki article are exaggerated IMHO.

I don't get the issue with the difference between read and write speeds though. Hard drives are slower ar writing too, usually about 2/3 the speed of a read for sequential access - all the SSD's I've used have a similar discrepancy (thanks to clever controllers and TRIM making writes far faster).

As to the longevity issue, I guess the jury is still out on that as these things have only been in the wild for a couple of years. But IIRC Intel said you could write 1/3 of the capacity to their drives every day for five years before you start to run out of writable blocks, and even then it's meant to fail gracefully.

I don't understand the point about SATA-based SSD's being slower for writes either; even the crappiest SSD I've used (an OCZ Agility) has random 4k writes somewhere in the order of 8-10 times faster than a 10k SAS drive. Smells like baldercrap to me but I'll go give the sources a read - here's looking forward to your edit!

Re:Not so fast.

[Hurricane78]

- Limited lifetimes
Funny, that's considered the major downside to platter drives. Anyway, this is the same as the first point.

NO. It’s NOT the first point. The first point is a symptom of this.
Funny? Not so funny when you notice that your music, movies, project backups etc, are all corrupted and lost. Also not so funny that you will notice it way too late, and all your backups will already contain partially corrupted files.
Actually there were viruses who did that exact thing: Write a bit of corruption here, a bit there, slowly, so it ends up on the backup media. If you ever had something like that: It’s the worst possible thing that can happen to your data.
And other than ZFS’s constant scrubbing ECC with checksums and a 3-disk mirror, nothing protects you from it. No RAID, no backups, no checksums alone.

- Performance of SSDs degrades with use.
Solved. See TRIM.

No. TRIM does not solve that problem. Do you even know what TRIM is?

Note that this is also a problem on platter drives.

I notice a pattern in your arguments: You say that there is the same problem with hard drives, and use the false dichotomy that there would only be the two states of “problem” or “no problem”. When in reality, the question is how much of a problem it is. And the answer is, that it’s much worse for SSDs.

- SATA-based SSDs generally exhibit much slower write speeds.
This one doesn't even make sense.

SATA limits the speed. Not only the bus. But also the internal controller. Simple as that.

Re:Not so fast.

[Anonymous Coward]

You merely saying "hard disks have these problems too" doesn't make it so. Where are your references for hard disks doing wear levelling, or their "performance degrading with use"? I'm not aware of either of those being true.

[citation needed]

Re:Not so fast.

[anethema]

You're just wrong on so many issues.

As for lifetimes, unless this is a VERY heavily used drive (100s of gb per day) the drive will far outlast any spinning platter drive. Even writing a 60gb per day, which is far more than I've EVER averaged, you're looking at 54 years before you run out of writes. Even taking into account the write amplification etc, this drive is going to last MUCH longer than any spinning platter drive.

TRIM absolutely solves the problem of lowered performance on SSDs with use. The reason performance degrades is because once the entire drive has been written do, the ssd has to do read, erase, write for every write instead of just a write. It doesn't know the deleted entry in the FAT actually means the data is no longer needed. TRIM tells it this. If you look at reviews, drives stay within a percent or two of their brand new performance with TRIM enabled.

His arguments are still valid about platter drives as well. Asymmetric read-write performance is a fact on SSDs, but even on their worst worst day, it is orders of magnitude faster than a spinning platter drive. So no, I wouldn't say the problem is worse on a SSD.

SATA does not limit write speeds until you hit the bandwidth limit of SATA. It does not pick on write speeds either.

Basically except for price per GB, SATA blows spinning platter drives out of the water in every way. They are vastly more reliable, quieter, less power, tougher(more rugged), have longer lifetimes, and fail very gracefully at the end of their life. If they were par in price/GB there would be absolutely no reason at all to use a spinning platter HD.

Re:Not so fast.

[slimjim8094]

- Wear leveling used on flash-based SSDs...
Oops! Wear leveling is done on HDDs too. And it isn't a disadvantage: it is a solution.

Wear-leveling is done for bad sectors on a HDD, not as standard practice.

- More expensive, lower capacity
Don't need to address that, since that is the topic of the article...

- Asymmetric read vs. write performance
Oops! Platter drives have this problem too!

Wrong. I just ran a benchmark and saw widely varying performance based on sector size and sequential vs random, but the reads and writes were the same speeds. I tested all 3 of my HDDs. If this is true, [citation needed]

- Requires TRIM
Solved.

- Limited lifetimes
Funny, that's considered the major downside to platter drives. Anyway, this is the same as the first point.

- Performance of SSDs degrades with use.
Solved. See TRIM. Note that this is also a problem on platter drives.

These two are related. SSDs may last longer than HDDs for certain use environments (space?) and types (maybe even a typical user) but they're definitely not a given. Try to defrag a MLC drive a few times and it'll be dead in a week (yes I know you don't need to defrag a SSD, but there are processes that can mimic it). And every HDD I've owned (dozens) for the past 15 years is still spinning, though I suppose I may be lucky. And a SSD will have cells go bad as a consequence of the technology, while a bad sector on a HDD is a fault. If a HDD goes bad, it's due to mechanical failure of the supporting systems, not degradation of the media itself.

--

Look, SSDs are great. I love coding on one because compiles are wicked fast. But they are not ready to replace HDDs. And they're not the panacea you're making them out to be. Like literally everything else, you pick the right tool for the job. I'll keep my backups and my movies and music and 11 days of recorded TV on my 2TBs of spinning media, and my OS, applications, and code on a smaller 128GB SSD.

Re:Not so fast.

[xtracto]

if I hadn't already spammed this thread with posts I'd have modded you up - other than price/GB, most of the problems in the wiki article are exaggerated IMHO.

It is something I have noted about wikipedia for some time, "editors" feel the need to add a bunch of "negative" or "positive" things in a wiki article just to make it view "neutral". Even if they do not know what they are talking about, or if it is completely false or FUD.

I think the last time I saw it was when reading about some homebrew for a console (they added "bad things" even though it is just the typical crap that console makers tell to scare people [oh noes putting a chip will make your Xbox explode!!])

Re:Not so fast.

[MobyDisk]

Funny? Not so funny when you notice that your music, movies, project backups etc, are all corrupted and lost.

Cute alarmist anecdote, but that is not how SSDs fail.

No. TRIM does not solve that problem. Do you even know what TRIM is?

Yes. TRIM was created to address this problem. The articles comparing TRIM drives to non-TRIM drives show that the problem has been solved. For references on this, just check out prior Slashdot discussions. Tom's hardware, storage review, and Anandtech have articles on this.

If you have evidence that TRIM does not solve the problem, please post it. Asking "Do you even know what TRIM is" doesn't provide any evidence that it doesn't work.

And the answer is, that it’s much worse for SSDs.

Actually, SSDs fail gracefully, while hard drives die suddenly with a crunching sound and a destruction of data. If an SSD runs out of sectors to write to - it becomes read-only and you have to buy a new one and copy your files over.

SATA limits the speed. Not only the bus. But also the internal controller. Simple as that.

I see that I didn't explain why this part of the article didn't make sense. The reason is because it argues against SATA, but the article is about SSDs.

Re:Not so fast.

[MobyDisk]

Each sector of a hard drive has a checksum. That checksum is used to verify that the data is in tact. During a write cycle, the drive reads back the values and verifies it against the checksum. If the checksum fails a certain number of times, the drive will map the sector to another physical location on the disk. You can probably find info about this by looking up S.M.A.R.T which allows you to query things like how many sectors have been reallocated as a result of bad checksums.

To combat this, Western Digital has something they call Preemptive wear leveling (more info here). I think other each drive manufacturer has their own way to deal with it.

Re:Not so fast.

[MobyDisk]

I have citations for everything you asked for, and then a few more:

Wear-leveling is done for bad sectors on a HDD, not as standard practice.

1) I still was considering that wear-leveling. Just after-the-fact wear leveling. :-)
2) However, Western Digital drives do have "Preemptive wear leveling" as a standard practice.

Wrong. I just ran a benchmark and...

On a HDD, writes are followed by an immediate read to verify the data wrote correctly. Take a look at some of the benchmarks on storagereview.com Compare the random writes to random reads: the writes are always slower. StorageReview doesn't compare continuous writes, because that is rare. However, you will find that there are drives specially designed for continuous writes for A/V purposes, that are made to address this issue.

Try to defrag a MLC drive a few times and it'll be dead in a week

My turn. [citation needed]. I'll provide my own references.
According to Kingston

For USB Flash drives, Toshiba calculated that a 10,000 write cycle endurance would enable
customers to “completely write and erase the entire contents once per day for 27 years well beyond the life of the hardware.”

Intel was more conservative with their report

...the X-18/25 SSDs have a mean time before failure (MTBF) rating of 1.2 million hours, which is on par with modern server hard drives. In addition, he claimed that the drives can withstand a workload of 100 GB worth of writes a day for five years.

They then go on to explain how it was actually hard to even write 100GB per day.

not degradation of the media itself.

Magnetic media does degrade. For one thing, it warps over time. I can't state all of the methods of failure.

Look, SSDs are great... But they are not ready to replace HDDs.

I agree. Just not for some of the reasons the Wikipedia article lists.

Why does it matter?

[SanityInAnarchy]

Encrypt it to begin with, especially if you've got a home media server. Store the key on a separate device -- say, a USB key. If you need to nuke everything, simply take some thermite to that USB key.

Probably overkill, but the point is, it's a lot easier to destroy a key than it is to destroy the data.

Spinning drives are already obsolete

[Weaselmancer]

There is something better. In every measurable way SSD is better. Better power consumption, better shock/drop resistance, lower heat, no noise, near zero seek times, they already can run SATA3, and longer lifetimes. Yes - longer lifetimes. Check this guy's math and you'll see what I mean.

The only thing that keeps them off the mainstream currently is price, which Moore's law will quickly fix.

I've always been a slow adopter of new technology. Hell, in high school I thought CDs were a passing fad. I can admit that. But this though...this one is easy. I don't think I've ever seen a technology upgrade that so completely outshined its predecessor in my life.

I agree that spinning disks will have their place for a while, much in the same way ice houses were still needed when the first refrigerators were being made. Only necessary during the transition to the better tech. Likewise, if the price were to match rotating media dollar for dollar today - you'd never see a spinning hd ever again, except in legacy systems.

Re:Spinning drives are already obsolete

[Tridus]

Storage space?

I can get a 2TB hard drive, right now. 2TB of SSD storage is unrealistically expensive, if you can even find a way to get that much in one system.

Software, games, and media are ever increasing in size. Drastically reducing the size of drives is not realistic in peoples computers.

Re:Spinning drives are already obsolete

[petermgreen]

2TB of SSD storage is unrealistically expensive
Agreed.

if you can even find a way to get that much in one system.
You can get a 512GB 2.5 inch 9.5mm high SSD now. You can stick two of those in one 3.5 inch bay so putting four of them in should be possible in almost any case. I'm quite sure that if one of the major vendors tried they could make a 2TB 3.5 inch SSD.

Software, games, and media are ever increasing in size.
While true to some extent in my experiance most PCs sold to normal lusers are still sold with drives of 320GB-500GB and this is more than enough for most people (with the exception of those who like to keep movie collections on thier computers but afaict most people don't).

Drastically reducing the size of drives is not realistic in peoples computers.
Some SSDs that are as big as typical HDDs in consumer machines are already on the market. That leaves two things to happen, firstly SSDs of acceptable size need to come down a lot in price, secondly someone needs to convince users that SSDs are worth it.

Re:In 5 years

[Renegrade]

You know, I've been meaning to get a CF-to-IDE adapter and plug that into my old Amiga 1200.

You see, it didn't spontaneously write to the OS disk. EVER.

The only times writes occurred there is if you saved prefs, copied files there, or specified it as a location for a file that you're saving.

I think later third party extensions caused writes to be a bit more frequent, and I don't know about the successor OSes (>3.1), but 3.1 and earlier and earlier had this .. read purity thing going on. I'd love to run it on an old one gig CF card, as that would be tons of space for 3.1. Turns out, many years later, that this behavior has some very serious benefits -- it's the perfect OS to run on a flash card.

Hell, the original machine only had 120 -megabytes- of space anyways when it was a primary machine, with like a 20 meg OS partition.

So, for some OSes, it would be the perfect media.

Re:In 5 years

[mrsteveman1]

These were all Seagate and WD drives from retailers

Re:In 5 years

[Anonymous Coward]

Mod parent down for being simply wrong. Power consumption is directly proportional to clock frequency, not the square of the clock frequency.

An input to a CMOS gate can be approximated as a capacitor, so each time the capacitor is discharged, an energy is consumed equal to the energy stored in the capacitor. The energy in the capacitor is 1/2*C*V^2 where C is the effective capacitance, and V is the supply voltage. The total power consumption is n*k*f*C*V^2 where f is the clock frequency, n is the number of gates and k is the activity level which describes the number of times per clock cycle each gate will change at its input (on average). The 1/2 gets absorbed into the k.

If you double n (two cores), but halve f, the power consumption doesn't change.

Re:In 5 years

[MrNemesis]

The point of the shift to 4k sectors (e.g. the WD "Advanced Format" drives) is that the amount of space needed for error correction at ever increasing densities was entering into the bounds of diminishing returns. Larger blocks mean less error correction is needed and thus more storage space for a given platter density. Anand has a pretty good writeup on it here: http://www.anandtech.com/printarticle.aspx?i=3691

Re:In 5 years

[RobKow]

P=I^2 R

So, for the same resistance, the heat is proportional to the square of the current.

Re:In 5 years

[K. S. Kyosuke]

On the other hand, you only need to install Windows ME, and you promptly get a fanless system.

Re:In 5 years

[aix tom]

Here, too.
My basic "swap cycle" for hard drives was

1) Buy them
2) Use as data storage 2-3 Years
3) Use as OS drive 2-3 Years
4) Use for swap space 2-3 Years
5) Throw them out

I have gone through maybe 25-30 drives for various boxes at home so far, and exactly ONE has failed me so far, while it was already on "swap space" duty. Usually the ones I throw out are about 8-10 years old, just because they are now even to small to be useful as swap space.

Re:In 5 years

[Al's Hat]

Tape also has the advantage of weighing significantly less than drives which is a factor to be considered when storage off site is required.

HDDs are Done When Google Says They Are Done

[FrozenGeek]

Seriously. Google is (believed to be) the largest single user of consumer hard drives. When they start replacing hard drives with SSDs, I will consider HDDs to be done. I wonder what price differential the power savings (don't forget the power for cooling) will cover?

Re:HDDs are Done When Google Says They Are Done

[MobyDisk]

This article claims Google is using Intel SSDs. There's no source though, and Google declines to comment. Oh well.

Re:HDDs are Done When Google Says They Are Done

[FrozenGeek]

Would not surprise me if they were using them on some of the very critical loci of their system. And I'd imagine that they would want to actually test the effects of converting to SSD so that they have useful data when trying to determine when to start switching.
A few test installations does not constitute a wholesale change in direction, but they do serve as portents of the future.

Re:HDDs are Done When Google Says They Are Done

[mooglez]

Seriously. Google is (believed to be) the largest single user of consumer hard drives. When they start replacing hard drives with SSDs, I will consider HDDs to be done. I wonder what price differential the power savings (don't forget the power for cooling) will cover?

They started planning that, 2 years ago?

http://www.informationweek.com/news/storage/systems/showArticle.jhtml?articleID=207602745

Re:HDDs are Done When Google Says They Are Done

[Hurricane78]

But then you will also have to use the same amount of redundancy that they use.
From a paper I know that they use 3 disk mirroring. Just like airplanes and spacecraft have everything thrice. Because with two disks, you won’t know which of them is right.

Re:In 5 years

[jim_v2000]

And everyone will wonder why we still call them "disks".

60 Gig = 127

[crazyfrenchmen]

Actually, today at newegg you get a 60Gig SATA II for 127 after MIR:
http://www.newegg.ca/Product/Product.aspx?Item=N82E16820227461

Re:In 5 years

[TubeSteak]

If you get rid of the fans, there won't be any funny/troll posts about Microsoft, Apple and Linux.

There were never really any fans:
They're all perpetual motion sterling engines running off the temperature differential of their own hot air and the chilling glares of everyone who thinks they're idiots.

Disks are dying -- AGAIN...

[sillivalley]

Pundits have been tolling the death knell of rotating storage for ... decades?

But somehow, the rotating storage business manages to innovate its way back to relevance -- Winchester technology, thin film heads, headerless architectures, increased spindle speeds, bigger caches, perpendicular recording, 4k sectors, continuing advances in encoding and ECC, continuing advances in media -- the advances keep coming.

And whatever happened to bubble memory, anyway? Wasn't that supposed to save the day and obsolete rotating storage once and for all? Isn't that what Intel promised us?

Re:Disks are dying -- AGAIN...

[1s44c]

It's not true until netcraft confirms it.

Re:Disks are dying -- AGAIN...

[bertok]

Pundits have been tolling the death knell of rotating storage for ... decades?

But somehow, the rotating storage business manages to innovate its way back to relevance -- Winchester technology, thin film heads, headerless architectures, increased spindle speeds, bigger caches, perpendicular recording, 4k sectors, continuing advances in encoding and ECC, continuing advances in media -- the advances keep coming.

And whatever happened to bubble memory, anyway? Wasn't that supposed to save the day and obsolete rotating storage once and for all? Isn't that what Intel promised us?

But despite all of those improvement, over decades, the minimum and average access speeds of drives hasn't improved significantly. A random read still takes milliseconds, which is a simply enormous time for a modern computer to wait. This in turn reduces the throughput in comparison to the streaming speed, making even the 'fastest' hard drives quite slow in practice for most workloads. I've seen benchmarks for 'enterprise' drives doing only a few hundred KB/sec for random reads. This is 2010! Computers can process data just a *tad* faster than that!

In comparison, most SSDs have random read times under a millisecond already, and some PCI-E versions are in the dozens of microseconds range. The random read throughput can be as high as 50% of the streaming throughput, and some drives are closer to 90%.

That's a night & day difference. It's like going from VHS to DVD.

And it's still not fast enough. A modern CPU can process IO data at rates of up to gigabytes per second, easily*. Physical drives have fallen woefully behind, and even SSDs aren't quite there yet unless you RAID a bunch of them.

*) For example, a SUN Thumper, which is just two average quad-core AMD Opterons, can send 1GB/sec (not gigabit, gigabyte) of iSCSI traffic down the wire. That includes reading from disk, decoding the ZFS structures, verifying the hashes, processing the iSCSI commands, and talking to the network card at 10Gbps. In comparison, the best consumer SSDs are still around the 250MB/sec mark.

Re:Disks are dying -- AGAIN...

[xtracto]

And whatever happened to bubble memory, anyway? Wasn't that supposed to save the day and obsolete rotating storage once and for all? Isn't that what Intel promised us?

I've never heard of that technology so I gave a quick read at Wikipedia article. Although the technology seems interesting, one of the wiki snippets at the end tells (me at least) why it is definitely a losing technology:

The Bubble System required a "warm-up" time of about 20 seconds (prompted by a timer on the screen when switched on) before the game was loaded, as bubble memory needs to be heated to around 30 to 40 C to operate properly.

Re:Disks are dying -- AGAIN...

[moonbender]

I don't know what happened to bubble memory, but I'm pretty sure there were no bubble memory products widely available from various manufacturers with massively better performance at fairly competitive prices. So maybe they're onto something this time...

Tape vs. Spindle

[itomato]

The price per GB is one concern, reliability and data transfer rate are two others. There are more - thermal considerations/power consumption, portability, media life (bit rot).

Most people have storage tiers - you can have fast/slow JBODs ready and waiting to accept and retrieve data, incorporating slower, offline tape which is SAN-connected, and managed by a robot, which can be transported via station wagon for great justice.

Price-wise, LTO-4 cartridges hold 800GB at a cost of around US$35, which also requires a minimum of one US$300 device to read/write the tape, and likely a dedicated connection on a dedicated interface (some flavor of SCSI), which may tack on another $100-$400.

I can go to Foo-Mart and buy 1TB of SATA for $100 or less - perhaps with it's own (slow by comparison) interface, enclosure and power supply.

Are they interchangeable as a solution? Nope.

Re:Tape vs. Spindle

[Leolo]

Where does one find an LTO-4 drive for 300$ A quick google reveals they are actually in the 1000-2000$ range.

Re:Tape vs. Spindle

[jafiwam]

1.5 TB drives have just been advertised this week for less than $100.

For another 50 I can get a SATA (eSATA/USB) cradle that lets me hot-swap drives, to add as many as I want.

Tape has it's uses, small time backup for lots of data is not one of them.

Re:Tape vs. Spindle

[steveg]

If you can find me a $300 LTO4 drive, give me an address. I'll jump on it.

Last I looked they were in the $2-3K range. They may have come down some, but to $300?

Re:In 5 years

[0123456]

P=I^2 R

So, for the same resistance, the heat is proportional to the square of the current.

You must have some seriously lousy wires going to your hard drives if you're losing more power getting it from the PSU to the disk than you are in the actual drive.

Ridding myself of the fans

[oscarwumpus]

I got rid of the fans and noise by not using my computers at all. i used my time machine to go back 3 days to mail this comment to /. so it would show up on this forum. As you can see, they get a lot of mailers doing the same thing so sometimes the posts aren't timed correctly.

Re:Ridding myself of the fans

[Miseph]

fr1st ps0t!!!

Can't wait to see if I actually got it in 3 months or so!

Re:Ridding myself of the fans

[operagost]

I tried going back to 1885 and sending a telegram, but it got here too early. Ironically, it ended up getting modded -1, Redundant.

Re:In 5 years

[dgatwood]

Particularly when the silicon runs on a fraction of that voltage, so you have all that conversion loss at the far end.

In fact, IIRC, it's the exact opposite. Most electronics like CPUs and RAM run cooler on lower voltage (at least up until the point where they start to misbehave at the desired clock speed).

Re:In 5 years

[Aeros]

magic? or wishful thinking at least...

Re:In 5 years

[dgatwood]

I have drives that are still running after 12 years in continuous use. I also have drives that were replaced after as little as 9 months because they failed suddenly (spin up and right back down, make grinding noises, or won't spin up at all). I've experienced both extremes from most of the major hard drive vendors (every vendor I've bought drives from). There's a wide degree of variation between drive models when it comes to how long they last, and sometimes even wide discrepancies between batches of a given drive model. Some drives are even predictable as far as how long they will last (within +/- tens of hours). Those are the scary ones.

The only safe hard drive is one containing the fifth copy of your data, buried in a lead-lined concrete vault in the middle of a Utah salt flat with armed guards, guard dogs, razor wire, and no electrical connection to the outside world.

Re:In 5 years

[dgatwood]

The thing is, SSDs are expensive primarily because of economies of scale. If a lot of major manufacturers started making these drives available for $100 bucks as a feature, you'd see much wider adoption, and it would be more profitable to build additional fabrication plants that would bring the costs down immensely. Even now, 32 GB USB flash sticks cost on the order of $70. SSDs cost twice as much solely because the controllers are made in such limited quantities.

Re:In 5 years

[CohibaVancouver]

I've never seen a consumer hard drive last even 3 years

Where are you using these beasties? In a sandstorm? Underwater? I've got two computers at home, one's about 9 years old, the other one is around 5 years old - No issues with the hard drives at all, no clicking, no "bad data"...

But they told me the cloud is the computer now.

[fastbiker]

Isn't everybody supposed to be storing everything in The Cloud soon? Capacity won't matter and it will be available all the time and it will just work over the all encompassing, ever reliable, internet? Web apps for everybody!

maybe

[JackSpratts]

what's behind the +TB drive race is home media and consumer servers don't need anything like the speed advantages of chip based storage. spinning platters easily serve content throughout the house where size/cost and the ability to hold giant files is the determining purchase factor. i paid $65 for an external 1TB drive last fall because i had to have something at least that big for my content, and 2TBs under $150 are now available. meanwhile an acquaintance just finished design work on asml's next generation chip fabber and if i can't reveal the specific sizes i can report they're shockingly small. however i just don't see silicon storage beating the price/density advantages of platters anytime soon, regardless of the author's predictions.

cheap and huge. when it comes to multi terabytes that's what most families need.

- js.

Re:Tiered Storage - Software joining SSD+HDs?

[thms]

Right, newer distributions mount /tmp as tmpfs as well. No such thing on my ancient install, so I use /dev/shm. That defaults to half the available RAM, which I think is more than the default for /tmp.
Talking about RAM for file storage: Being able to just tell the OS to keep a certain file cached no matter what the scheduler says would be nice too (wasn't the sticky bit on files used for that a long time ago?).

As for cluttering, /tmp gets cleaned upon reboot (that doesn't happen anymore?), and since I shut down for the night...

Re:In 5 years

[PRMan]

Pity the lesson of Y2K went unheeded - where every COBOL programmer was paid whatever they asked to fix their code, but after should have all been taken out to a field and shot in the head.

You don't remember the days of limited storage, do you? Those 2 extra bytes times 100000 records * 20 date fields was 1/10 of your drive back then.

Now get off my lawn!

Re:In 5 years

[UnknowingFool]

where every COBOL programmer was paid whatever they asked to fix their code

The initial assumption here is that there was a design flaw in their code. It wasn't a design flaw; the code was simply never designed to be running for this long. In some cases of very old code, it wasn't practical to use a 4 digit date when the code was written. In some cases the programmers warned well in advance that it would need to be fixed but that costs money and business don't willy nilly spend money unless they have to spend money.

Re:In 5 years

[bluefoxlucid]

No, the power comes in from the wall to the PSU, where it gets rectified and buffered, and regulated. If you pass 15 amps through the regulator, it overheats. Take a 25W resistor and pass a 10 amp current through it and it gets VERY hot; pass a 1 amp current at 10 times the voltage, and it doesn't.

Re:In 5 years

[John Napkintosh]

With the exception of optical media, didn't we collectively drop the word 'disk' in the 90s? I thought they were just "hard drives", and now "solid state drives" now.

I'll still use the acronym HDD, but I'll never actually say "hard disk drive".

Re:In 5 years

[b4dc0d3r]

They both go around in a circle - one is horizontal and one is vertical.

Oops, I missed the word "price". Vertical should be more expensive due to the varying g-forces. Horizontal spinning does not have to counteract gravity.

More space!!

[Humus B. Chittenbee]

My son - admittedly a music major - has over 100G of music on his laptop ... from an extensive classical and opera collection to death metal (of many varieties) and jazz (his particular interest) and every other possibility [except I do not think he has any country/western.] I am a lightweight ... I only have a little over 20G. ;-)

Re:In 5 years

[Hognoxious]

Now, there's some point where drives become "big enough"; my disk space growth has slowed a bit from what it was a few years ago; I've basically only roughly doubled my use over the last 5 years.

Just wait till HD pr0n becomes commonplace.

Re:In 5 years

[bluefoxlucid]

You apparently don't know how DDR works or why it's still a 100MHz clock internally with an 8 byte buffer rather than an 800MHz clock. Lots of components run at a low clock rate with a buffer and multiplier these days, or run in massive parallel at low clock rates (GPUs).

Re:In 5 years

[mcgrew]

Oh, the ignorance of youth. Computers were over fifty years old in Y2K, and your cell phone is more powerful a computer than any built in the fifties. Hell, a Hallmark card is more powerful and sophisticated. They used two digits for years because they had to. There simply wasn't enough data storage (which oddly makes this otherwise offtopic comment on topic). You take your terrabyte disk drives and your gigabyte SSDs for granted, but early system were measured in kilobytes.

An example is the IBM 1401 that was announced by IBM on October 5, 1959 (I was seven years old at the time).

The 1401 was available in six memory configurations: 1.4K,[4] 2K, 4K, 8K, 12K, or 16K (a very small number of 1401s were expanded to 32K by special RPQ - Request for Price Quotation). An optional "Advanced Programming Option" allowed for additional flags for 3 characters within the first 100.

Legacy data and cheapassed managers kept the two digit dates around, and programmers and systems analysts warned management of the coming doom, but were ignored until it was almost too late.

A COBOL programmer in the 1950s would be dumbstruck by what we have today. Actually, I'm dumstruck as well; cell phones, flat screen computers, and self-opening doors in Star Trek were impossible; science fiction. You young folks can't imagine how primitive things were when I was a kid, and nobody dreamed we'd see SSDs.

Re:In 5 years

[DIplomatic]

Interesting. Hard drives replace tape backup. GOOGLE replace hard drives.

There, I fixed that for you.

Re:Tiered Storage - Software joining SSD+HDs?

[Chris Burke]

Also, you could overrule the prediction algorithms easily to cache the music you want to listen to or the database you are working on.

I'd overrule the algorithm to not cache my music files (on the SSD) and instead be read directly from the HDD, since the transfer rate is negligible and startup latency is not a big issue. Same with movie files. HDDs are just fine for streaming. It's frequent, latency-sensitive random access patterns where the SSD cache would pay dividends.

Other than that, I like your proposal. Make it so!

Re:In 5 years

[afidel]

SLC flash has essentially unlimited writes for all but the most corner of corner cases but it's incredibly expensive on a $/GB basis (though it's much cheaper TCO than short stroking a bunch of 15k HDD's on a $/IOP basis).

Why can't we have multiple reading head ...

[kai6novice]

Why can't we have multiple reading head spinning media to increase access speed and reduce spinning = reduce energy cost & reduce the chance of damage. For example, instead of a single harddrive arm, we have 2 or 3, same for optical device.

Good spinning media is in high demand.

[bregmata]

My wife has plenty of spinning media. She has two wheels and needs a constant supply of media to spin into yarn. With the loom and plenty of knitting projects on the go, there is never a shortage of demand.

Re:In 5 years

[TheThiefMaster]

However lower clock frequencies also require less voltage to keep a stable signal (and correspondingly, high frequencies need to be driven by a higher voltage). Taking this into account, power use does drop much more than linearly when clock frequency drops.

Re:In 5 years

[yacc143]

Well, I'll question the more reliable part. Despite having owned way more harddiscs in the last decade that I've owned even tape media (tape was a backup solution only for some years), I had more unreadable tapes than unreadable hdds.

Again

[Weaselmancer]

The only thing that keeps them off the mainstream currently is price, which Moore's law will quickly fix.

Likewise, if the price were to match rotating media dollar for dollar today - you'd never see a spinning hd ever again, except in legacy systems.

Re:Again

[jedidiah]

...except drives aren't going to go below a certain price.

This is true of stuff in general. You get to a certain bottom
point where the market resists selling you anything cheaper.
So you end up with a certain certain price that remains fixed
over time while capacity increases.

Unless SSD's get a lot bigger, they aren't going to be price competitive with spinny disks.

Cheapness alone isn't going to do it.Although time will tell.

Re:In 5 years

[edmudama]

actually it's physics...

The unique failure mode for SSDs is related to the NAND substrate wearing out. However, modern SSDs detect this and rewrite their data before it's at risk of being uncorrectable. While a brand new drive may hold data for 100 years if you've only written it once, a drive at the end of warranty having been used to the max may only hold data in your closet for a year before it starts showing errors on readback. Just watch your SMART tables and you should be fine, since SSDs tell you how many write cycles are remaining as a percentage.

The opposite of that is the failure mode of most hard drives which is a head crash either due to mechanical damage or ESD. Once that happens, your entire drive is dead, unless you have many thousands of dollars for a recovery service.

The above assumes that the chance of failure due to things like PCBA solder failures etc. is roughly common and similar across both types of products.

Re:In 5 years

[commodore64_love]

I like how you think.

Last year I tried to slowdown my Pentium 4 from 3000 megahertz, because I got tired of hearing the fan, but Intel apparently didn't include that capability. I don't need 3000 megahertz just to surf the web or watch a youtube video. Even my lowly 600 megahertz laptop is fast enough to do those things, so being able to slow the P4 clock would be a nice feature.

Back to topic -

Does this mean Nintendo will return to using carts?
Like on my old N64? I like carts - virtually indestructible.
And with SSD they'll be rewritable.

Re:In 5 years

[LaughingCoder]

Well, arguably, GOOGLE (or, more precisely, "the cloud", replaces both tape and hard drive as far as the end user is concerned. Of course my point still holds when discussing what Google (i.e. the cloud) uses for online storage and backup.

Re:In 5 years

[bsane]

Which will be true for the cables to the disk, but neither a traditional HD or SSD are resistive loads.

Re:In 5 years

[Carnildo]

Interesting claim, considering that these drives don't meet any of my needs. My computers fall into two categories: those that need 1-2GB of disk space, and those that need multiple terabytes of disk space. For the former, I can stick a $15 CompactFlash card in a $5 adapter; for the latter, spinning disks are still over an order of magnitude cheaper for the same capacity.

I don't expect this situation to change in the next five years.

Re:In 5 years

[commodore64_love]

>>>get a CF-to-IDE adapter and plug that into my old Amiga 1200..... it's the perfect OS to run on a flash card.

I'd just buy a whole new Amiga with OS 4.0. You can get them off Ebay, or places like Great Valley Products (GVP), and with a processor fast enough to handle modern video (like youtube for example) - 1000 megahertz. As I recently discovered that old 68020 CPU is simply too slow to be useful in the modern world.

And AmigaOS still has the advantage of simply flipping the switch to turn it off. And a startup time of less than 10 seconds.

mainstream use of HDs could end early next year

[Tumbleweed]

MOST people (no, not those of us here on Slashdot) don't even need an 80GB drive these days for their computer. OS, apps, and downloaded music generally fits really easily on an 80GB drive. Most people never rip a movie or anything like that. I just got a 64GB SSD to use as a boot drive, from Newegg for $144. With the coming die shrink for flash memory coming in Q4 this year, that price could easily be what you pay for a 128GB drive for the coming holiday season, or early next year, which is all most people will need.

Re:In 5 years

[mswhippingboy]

Yea sure, tell that to the pile of drives in the corner of my office that gave up the ghost in 1 to 2 years.....

Re:In 5 years

[tomhudson]

The only safe hard drive is one containing the fifth copy of your data, buried in a lead-lined concrete vault in the middle of a Utah salt flat

You are infringing on our IP. Pay your $699 burial-in-a-salt-flat license fee.

-- SCO

(to be in Chapter 7 by the end of April, 2010)

End of HDDs? Not likely soon.

[TheGreatOrangePeel]

Q: "SSD Price Drops Signaling End of Spinning Media?" A: No

Let's face it, a hard drive to hard drive is currently the backup method of choice. Anyone who denies it can be pointed to a plethora of, "Ask Slashdot: How do I store my data?" discussions. Just like when tape drives could store more than the systems hard disk, a hard disk offers to hold more than the average SSD. Never mind the fact that when an SSD fails, it's more than likely end-game for your data. But when a HDD fails, there's any number of data recovery companies at hand to restore it.

The introduction of SSDs will add pep to the computers we use, but hard drives will continue to be the workhorse for storing the bulk of our data for a long while to come.

Re:End of HDDs? Not likely soon.

[cffrost]

Never mind the fact that when an SSD fails, it's more than likely end-game for your data.

I've been under the impression that SSDs' most common failure mode was to enter a read-only state. Do you know this to be incorrect?

Re:In 5 years

[mswhippingboy]

Software is different. I still hear stories of folks running autocoder programs!
Hardware will most definitely eventually die, and when it does, it will get replaced by the most cost effective hardware available.

Re:In 5 years

[Skratchez]

Oh you younglings. Back when I worked with Charles Babbage all we had was brass and steam. Now all the dorks on the block are wearing aviator goggles and leather and calling themselves punks. Disgraceful behaviour. Just disgraceful.

Re:In 5 years

[bsane]

You can't use resistors as your model of computer components- well you can, but you get unreasonable conclusions like the one above.

Take as an example CMOS tech, theres a pretty good run down of why any type of semiconductor doesn't act at all like a resistor when it comes to power dissipation.

http://en.wikipedia.org/wiki/CMOS#Power:_switching_and_leakage

FTR- I'm not saying P=I^2 R is wrong... Its certainly correct, the challenge is coming up with 'R' when you're talking about switching semiconductors (or anything else really).

Re:In 5 years

[styrotech]

A COBOL programmer in the 1950s would be dumbstruck by what we have today. Actually, I'm dumstruck as well; cell phones, flat screen computers, and self-opening doors in Star Trek were impossible; science fiction.

Hehe, in the 50's just having TV shows like Star Trek probably seemed like impossible science fiction :)

Re:In 5 years

[tomhudson]

Everything is essentially analog when it comes back to the physical world.

My fingers are all digital, you insensitive clod!

Great for digital calculations.

And raising one of the middle ones signals "sign-extended mode".

You can have my digits when you pry them from my cold dead hands! :-)

Re:In 5 years

[gtbritishskull]

But, there is more to it than just list price. If the disks use less energy, then that can add up quickly, especially at big server farms. A drive you have to replace half as often is more than worth twice as much, because you are at a lower risk of data loss and don't have to expend the labor to replace it.

Reliability of NAND flash

[droopycom]

I read NAND flash datasheet as part of my job, and I'm really wondering what those guys are thinking and what it means for SSD.

There are two things I look at in the datasheet:
- Data Retention
- Program/Erase cycles.

Data Retention is how long the data on a flash will last before you need to refresh it.
Program/Erase cycles is how many times you can re-program a flash sector before it stops working.

It seems that, as NAND flash get bigger capacity, and as they move from SLC to MLC technology, the shorter data retention get, and the smaller the number of P/E cycles get.

Right now, I see 5000 P/E cycles and 5 years data retention. Roadmaps show both of these getting worse.

Now, 5 years data retention is probably still okay for most use where you dont keep devices more than 5 years. But 5000 P/E cycles seems very low for a PC or even smartphones. I would think that wear leveling algorithm can only do so much....

Re:Reliability of NAND flash

[Rockoon]

Now, 5 years data retention is probably still okay for most use where you dont keep devices more than 5 years. But 5000 P/E cycles seems very low for a PC or even smartphones. I would think that wear leveling algorithm can only do so much....

Even on the smallest drives..

5000 write cycles * 40 gigabytes = 200 TB of writable lifetime, At 5 years that 40 TB per year, 3.33 TB per month, 109 GB per day. Lets assume that the wear leveling algorithms actually suck complete fucking nuts, such that the actual limit is only 50 GB per day for 5 years.

Now check your own usage over time. Seriously. Do it. Stop making stupid assumptions. On windows the program you can run Resource Monitor, so no excuses if you are a windows boy. Do it.

It takes a special kind of user to be burning through 50 GB per day, every day, for 5 years. This very same special kind of user burns through traditional HDD's very quickly as well. Quite simply, if you do not frequently have to delete things to make room for new things, you are nowhere close to 50 GB per day. Check your own usage. Seriously. Do it. Stop making stupid assumptions.

I really hate people so bad at math that they dont even bother to do the simple stuff. A fucking 4 year old could have told you that you were wrong.

Re:Reliability of NAND flash

[droopycom]

Wow, indeed that made you real made... I'm sorry, I was thinking with my ass, (but thats slashdot)..

In the meantime I found some links to some study by Toshiba:
http://silvertonconsulting.com/blog/2010/01/08/toshiba-studies-laptop-write-rates-confirming-ssd-longevity/

The Toshiba study contains two main claims:
- How many GB/day their drive can support (22GB/day on their 64GB model)
- Average usage for laptop users (9.2GB per day for heavy user)

Now, those are laptop users (presumably windows). And these are average numbers. There are no numbers in a server environment. I could not find what they consider a heavy user, or what the variance was on those numbers. Can your 4 year old find those numbers ?

What configuration where those laptops running? Dont know... What happen if I have a laptop with no such much DRAM, but I use a big swap file to compensate. How heavy is the use of the swap file ? Can your 4 year old tell me ?

What about write amplification? What happen if I use a Filesystem with smaller blocks ? Can your 4 year old tell me ?

The toshiba study shows that just enabling hibernate and auto-save double the amount of data written. What other feature may double that ? Can your 4 year old tell me ?

Btw, WD has their estimate at 17.5GB/day for their 64GB model. If this was a 40GB like your example it would be 10.9GB/day, not 50GB, so obviously their wear leveling algorithm are not as good as your fucking 4 year old think they are...

Re:Reliability of NAND flash

[Rockoon]

My 4 year old read that the average user was only writing 2.4GB/day and said that someone who picked a theory and then tried to find supporting anecdotal evidence, but could only come up with extreme examples such as 9.2GB/day, is a rotting melon head.

How long do HDD's last under extreme usage? You seem to think that they last at least 5 years, but I know for a fact that an 8% yearly failure rate is a good approximation. The chance of a modern HDD surviving 5 years in a heavy load environment is 0.92^5, .. only 65%.

Welcome to the real world, dopey.

Re:In 5 years

[snooo53]

That would be great. No fans, no platters and, no noise. Almost every computer hardware failure I've had was because of a cheap/faulty stock fan or a dead hard drive.

Re:In 5 years

[guruevi]

Actually, back in the day, all we booted from was SSD (a few kB of ROM) because spinning media (floppy's and 'hard' drives) were freakin' expensive, not to mention gigantic and slow. This made stuff like instant-boot very normal to have back in the day. In the mean time we decayed to using ever faster spinning media until the hardware couldn't go any faster (15k drives since 1997) and the capacity couldn't increase (perpendicular recording a couple of years ago) and we waited minutes for our OS'es to boot. Now we're back at SSD's which don't really scale very well for large amounts of data (smaller chips means more expensive and more potential errors) until somebody finds a better format for storing large amounts of data cheaply (probably in the realm of 3D optical storage) which will slow us down again a bit.

Re:In 5 years

[Carnildo]

More significantly, long-term storage would run you $0.10 per byte per month. Those extra two bytes times 100,000 records times 20 date fields would run the bank a half-million dollars a year in increased data archiving costs.

Re:In 5 years

[TheRaven64]

Remember that flash drives still have a limit on the total number of writes; spinning drives don't

Seriously? Spinning drives don't have a limit on the total number of write? You actually believe that? What I think you mean is that spinning drives aren't guaranteed to work for at least a minimum number of writes.

Re:In 5 years

[pantherace]

Intel did not include this capability into it's desktop chips for a while, other than a temperature sensor triggered slowdown. (So, You could slow it down by removing that fan...) AMD introduced their cool'n'quiet for desktops, which clocked down, in the Socket 939 Athlon 64s (All Socket AM2 and AM3 procesors from AMD should have it), Intel put a similar thing into very late pentium 4s and core 2, and iX chips. Chances are your P4 was earlier than that, as the last few P4s revisions, only got to 2.8GHz if I recall correctly.

Cartridges are more expensive. My guess for that is that they won't.

Re:In 5 years

[AliasTheRoot]

I'm doing this already, it works bloody well. I have my OS and Programs installed on a 120GB SSD, which sits around 50% utilisation and use NTFS junctions (aka symlinks) to map storage for stuff that doesn't need superfast seek speed (aka data) onto a group of 1.5TB drives. It takes a little management, so isn't quite ready for the average user yet - but you do effectively get something like 5TB of online disk space combined with SSD performance.

Interestingly, i've found on Windows 7 that by running OS/Programs from the SSD that the contention issues you would normally get on a spinning disk are mitigated a great deal - and there's no noticeable hit with having the entire user profile (including junk like web cache). The system is booted in 5 seconds after finishing its POST, and the desktop is snappy right from the get go.

As to the hard disks as backup, it works pretty good. At work we have maybe 1.2TB in a full backup - we do a weekly full backup and incrementals onto LTO tapes, and a second weekly backup onto a consumer grade SATA 1.5TB drive in a USB cradle. The SATA drives are taken offsite in case the office burns down, £80 + carrying a few hundred grams around makes for really cheap and fast data transfer.

Re:In 5 years

[Pentium100]

Depends on the purpose, really. As the basis for your OS, the number of writes might be an issue, but for general user data it's less so.

This sounds kind of backward to me. The advantage of SSD over HDD is speed (especially seek time), but only the OS really benefits from reduced seek time, and what benefits the most is the pagefile, which gets written often. Only in certain circumstances user data would benefot from reduced seek time mostly video editing etc. Movie files not intended for editing, Office documents, audio files and photos won't benefit from reduced seek time, but SSDs will be more expensive per gigabyte than HDDs for some time.

So, it's more likely for one to have a small, but very fast SSD as a system drive and one or more slow, but big hard drives for data. This is almost what I use. System drive is 36GB 15000RPM HDD and files reside on a bunch of 5400-7200RPM 40-750GB hard drives. When I want to safely store files that I don't plan to access often I write them to LTO2 tape.

Re:In 5 years

[jedidiah]

Wake me when there's a 10TB SSD and no compelling reason buy a cheaper spinny disk.

Otherwise, you're just blowing a lot of smoke.

Re:In 5 years

[mcgrew]

Yes, since it first aired in 1966. But that's only six years after 1960, and there was a dearth of science fiction TV in the 1950s; I can't think of a single one. A COBOL programmer in the fifties would likely have watched Star Trek when it did come on.

Re:In 5 years

[profplump]

That's absurd. If you were trying to save space you could have used 16 bits to store unique dates for almost 180 years. With another 8-bits you could make that 46k years, and still used only half the space per date as the ASCII YYMMDD format.

So if they goal was simply to save space the programmers were retarded. The goal was obviously to make it dates easy to program around/display/etc.; saving space was at best a secondary goal after someone already decided to store the date as an ASCII string.

Re:In 5 years

[Rei]

Yep, that's what sector remapping is for. And at only $100, why *wouldn't* you want that kind of performance for your root partition? Sure, that's rather expensive for storing music or videos, but for a root partition, that's a great price.

Re:In 5 years

[TheRaven64]

It's also worth noting that the IBM 1401 and its sister, the 1620 (nicknamed CADET: Can't Add, Doesn't Even Try, because it implemented addition and multiplication via two lookup tables in core memory) used decimal characters, not octets. Each byte was 6 bits and stored either a value from 0-9 or a special code (negative word follows, end of word, and so on). Now, with octets, if you were very short on space you would store years as a 256-year range starting, say, in 1900 and going until 2156. This would cost you 8 bits per record. Back then, it cost 12 bits to store two decimal digits or 24 bits to store four. Given that core memory was around $1000/KB, using 24 bits per date entry was not a very good solution.

Re:In 5 years

[jimicus]

You don't remember the days of limited storage, do you? Those 2 extra bytes times 100000 records * 20 date fields was 1/10 of your drive back then.

Now get off my lawn!

Why on Earth would you store each digit of the year in its own byte?

Re:In 5 years

[Pentium100]

Take a 25W resistor and pass a 10 amp current through it and it gets VERY hot; pass a 1 amp current at 10 times the voltage, and it doesn't.

Actually, no.

Let's say I take a 1Ohm resistor and connect it to an ideal 1V source. The current will be 10A, and the dissipated power will be 1V*10A=10W.
Or, I can take a 10Ohm resistor and connect it to a 10V source. The current will be 1A and the power will be 10V*1A=10W.

In both cases the resistor will be at about the same temperature, assuming both resistors are the same size and shape.

Re:super cheap CRTs...

[Hatta]

I did add the qualifier "brand new".

Re:In 5 years

[jedidiah]

I've had drives last 7 years and still be usable even after they were taken out of service.

Quality of a particular device boils down to the manufacturer and whether or not they want to be bothered. This has squat to do with the particular technology. If a vendor wants to sell you a crap spinny disk then they will sell you the same kind of crap for an SSD.

There's plenty of computer parts without moving mechanical bits that give up the ghost in 2 years.

Some of it even has an Apple logos [ducks]...

Re:In 5 years

[MillionthMonkey]

Every week or so a van leaves Google crammed full of hard drives containing their current backup of the Internet.

Re:In 5 years

[petermgreen]

It used to be that saving conversion steps was a good thing both from an efficiancy and a cost point of view and so PSUs delivered at 5V which was what the logic of the day ran on. They also supplied a (relatively) small ammount of 12V for running drive motors.

When components started needing 3.3V initially they were run from local regulators but when ATX was introduced it added a dedicated rail for 3.3v.

However things changed again, firstly even lower voltages started to be used. Secondly chips started having seperate core and IO voltages with the core voltage often being highly chip specific (some chips even have multiple internal voltages).

These voltages could not be reasonably supplied from the PSU both because there are so many different ones and because the resistance of long cables would be such a big deal at those voltages (both in terms of power loss and in terms of voltage stability) so these voltages HAVE to be supplied by local regulation.

Now initially they used the 5V and 3.3V lines to derive lower voltages but it tends to be more efficiant overall to use a higher voltage for the PSU-local regulation link. Therefore from the pentium 4 onwards intel switched to using 12V (the highest voltage in use in standard PCs) for this link.

Now the same would be true for SSDs, afaict 2.5 inch SSDs like laptop HDDs currently run on 5V because that is the only voltage that can be relied on. This voltage almost certainly gets downconverted internally and moving to 12V as a source for that would probably improve efficiency a bit.

Having said that looking at the power consumption figures of a well known SSD series I don't think SSD power consumption is significant anyway even in a system with an atom processor.

Re:In 5 years

[ultranova]

this is what happens when you let free markets thrive, and don't bog them down in regulations and laws. You get nice steady improvements, and cheaper prices.

Now if only they lifted antitrust regulations, so that AMD and Intel could form a cartel and stop all this silly competition that's hindering technical progress!

Re:In 5 years

[jedidiah]

Perhaps you should stop taking advantage of the price drops that inevitably come when a particular make/model of drive becomes notoriously unreliable.

My only notable drive failures come from just that.

Re:In 5 years

[fullfactorial]

Yeah, if you really want to compare apples to apples, measure MTBF. Oh, and let's not forget the SSD's far superior ability to decay gracefully.

SSDs have a better MTBF, but I think you have the graceful decay backwards. Good SSDs do wear leveling and use SMART to tell you when your ten-thousandth write is approaching. But once they die, they're dead. Solid-state failures are a lot less predictable and more unforgiving than mechanical failures. (For reference, read up on the Poisson Process as it relates to solid-state failures.)

Re:In 5 years

[mcgrew]

I certainly hope you're going for "funny". It took almost as long for my TS-1000 to power up in 1982 than my new netbook takes; its processor was 1.4 mHz, the processor in my netbook is measured in gHz, but it has to load a LOT of data to boot up.

Re:In 5 years

[jedidiah]

My situation is pretty much the same. Except my "client" machines need 5G instead of just 2.

I am eagerly awaiting the first 10TB disk. I suspect it will be a spinny disk and not an SSD.

Re:In 5 years

[HiThere]

You are an idiot.

At the time most of that code was written, 32K (words) was a large computer. You SQUEEZED the bits into words tightly. TIGHTLY. People recommended tricks like XORING two pointers together to save space, at the cost of additional computation. And mainframe computer time was in the neighborhood of $700/hour. And that was before several rounds of 12% inflation. At that time a paperback book cost between $0.50 and $0.75, to help you calibrate what that meant.

Also turnaround for many programmers was once per day via courier.

At that time two digit years were the appropriate choice. Four digit years didn't become reasonable, by and large, until the 1980's or even later. (Remember when we moved from mainframes to CP/M computers, our disk storage was trimmed to around 70KB. And our RAM was limited at 64KB. It wasn't until personal computers got hard disks that this limit was lifted. (Networked hard disks came later for most people.)

So for anything written after 1990, you might well have a point, but that's not the code you're dissing. Idiot.

The other respondent who said you should have blamed the managers was more reasonable. Unfortunately current management theory claims that managers don't need to know anything about what they're managing. So the individual managers, themselves, probably aren't to blame. I'd put the blame on the general managers, who should know better than to accept that theory. (Though at their level it becomes true. But a part of their job is to know how the job requirements change as the degree of separation form the actual work increases, and they generally fall down on that. Badly.)

Re:In 5 years

[mv2s]

You didn't ask, but I will get off your lawn anyway...

Re:In 5 years

[MillionthMonkey]

Yeah. Good thing we didn't "bog down" SSD drives in "regulations and laws". Mandating rear view mirrors, maybe.

This isn't cheap

[Evro]

$100 for 40 gigs isn't really cheap. It's $2.50/GB. I bought a 128 GB SSD a year ago for $340 - not that much more. By contrast, you can get SATA storage for under $0.15/gig. I think most things will tend toward SSD as the price drops, but this doesn't seem like a price drop. When SSD hits $1.00/gig it'll probably take off. SATA will still be used for huge raids though for the simple fact that even at $1.00/gig it's 10x the price of SATA.

Re:In 5 years

[Pentium100]

My newest hard drive is 3 years old (7200RPM, 750GB). All others are older, and are on almost 24/7. For example, a 40GB Maxtor drive reports that it has been spinning for ~58k hours, so that's ~6.6 years. Here's a report from speedfan. Oh, and the UDMA error count has been like that for a few years now.

And here's another report. This 120GB hard drive has been spinning for ~39k hours.

The drives are much older than the hours they have been spinning because a few years ago I used to turn off my computer at night, so the "Power on hours" number is lower than the age of the drive. Now, both these drives are on almost 24/7, I also have some newer and bigger drives, but none are younger than 3 years old, becasue 3 years ago I stopped buying new hard drives. I use LTO1 and 2 tapes for stuff I don't access frequently so I don't need more hard drives.

Re:In 5 years

[Rei]

What's so forgiving about the magic white smoke getting out of a hard drive after a head crash?

Re:In 5 years

[spazdor]

I was under the impression that SSD's remain readable forever even after they become unwritable.

If I'm wrong about this then I stand corrected, but if I'm right then that is a highly desirable trait when you're worried about preserving data integrity across a hardware replacement.

Re:In 5 years

[HiThere]

Which keeps longer if you stick it on the shelf and forget about it?

If you're thinking about backup, you should be concerned about long periods of time. 20 years at the bare minimum. Reports are that DVDs don't last that long. Disks freeze up and need expensive repair to recover the data. How do SSDs stack up here. (Don't judge by current capacity, we're in the very early days yet.)

P.S.: *I* don't know. If you do, I'd like to hear your answer.

and robustness

[t0p]

The thing I like most about my EeePC is that I can drop it over and over and over again and it just keeps working. Do that to a laptop with HDD and you can kiss it goodbye. Yeah I know I'm a careless, clumsy klutz. But so are a great many other laptop/netbook users. The ability to treat your equipment like a shack of sit without fear of malfunction is priceless.

Re:In 5 years

[cbreaker]

Certainly, but your hard drives just sit around in one spot and never move. If tapes are properly cared for, they do last a long time, and you have to rotate them out of service if they're used a lot.

Tapes often get a raw deal. People snatch them from a drive, they sit them on top of the server, they don't put them back in their cases and they have no shielded storage location. They get full of dust, they never get retentioned...

But I'd like to see how long your hard drives last on a normal backup cycle of being transported back and forth between off and on site.

People always talk of the raw cost of the disk or tape. While a tape library system is not cheap, a storage array system is even more expensive. You can have a full cabinet sized tape library system with several drives for much less than a Clariion, Symm or EVA of similar capacity.

Many large enterprises use several tiers of storage, even just for the backup system. Where I am now, they have high end Tier 1 storage, lower-end Clariion FC Tier 2, Clariion SATA Tier 3, and for backups they use TSM Disk arrays (T1), VTL system (T2) and a massive tape library (T3 and Off-Site.)

The technologies aren't exclusive. They can compliment each other.

Re:In 5 years

[dgatwood]

Ah, but all this efficiency assumes that they'll be sane and will use switching regulators. Linear regulators are ten for a buck in quantities. :-D

Wow, does that make me cynical, or what?

You are right that running around low single digit volts isn't efficient. That said, I thought you were suggesting running the actual silicon on 12V, which would be way worse....

Stepping down to a volt or two at the drive level is perfectly fine. It certainly does mean that you can get away with smaller wires without incurring appreciable voltage drop. On the other hand, the current drain from these things is probably so low that I'm not sure if it will make enough difference to worry about. :-)

Re:In 5 years

[HiThere]

Captain Video. He had both spaceships and video phones.

I'm sure there were others, but most of the time we didn't have TV.

Killer app ?

[obarthelemy]

The issue is that SSDs are not really a revolution, just an evolution. A bit less power consumption, a bit less noise, a bit more performance, at the cost of mucho money, more performance variability, and an untested lifespan.

I personally am not in the market for an SSD yet. Gaining a handful of seconds every time I launch apps is irrelevant, since I no longer launch apps that much: Windows and Linux are stable enough to never require a reboot, so I start the PC, launch my apps... and they stay launched for a few days or weeks.

At work, I imagine there's plenty of tiered solutions already.

What would really make me think about it is some kind of ReadyBoost for SSDs, where the SSD acts as a cache for the HD. This way, even a smallish SSD (30 gigs ?) would provide a sizable performance gain. Right now, putting a whole OS or App on an SSD, while prolly only 1/3rd f their code gets read regularly, is majorly inefficient. Hopefully MS and Linux are working on it.

Re:In 5 years

[mako1138]

Take a 25W resistor and pass a 10 amp current through it and it gets VERY hot; pass a 1 amp current at 10 times the voltage, and it doesn't.

V = IR. You can't do what you have proposed, it's impossible. If you reduce the current by a factor of ten, the voltage must decrease by a factor of ten as well, the resistance being constant.

Re:In 5 years

[Idiomatick]

They have such computers. You can fill a computer with liquid and then use a bubble machine from a fish tank to keep it cool enough. That is if you are looking for performance. BUT people have also made no moving parts computers with low heat/power chips. And huge ass heatsinks. So it is totally doable.

Re:In 5 years

[aix tom]

Two reasons basically

1) Size. Swap needed the least size.

2) I didn't need swap maybe 99.99% of the time. Just for the odd over night batch run doing 3D rendering for example. It wouldn't have been too bad if batch run had failed because of the swap drive dying.

wake me up when

[JustNiz]

wake me up when you can buy a 1 TB SSD for like $200.
The reason I like spinning media is that it can do practically infinite number of read-write ops, while an SSD sector can only take about a million writes. I know there's a load-levelling algorithm, and also 1 million is too many to worry about anyway, but when you consider its use for swap it really might not be. Call me strange but I don't like the idea of my expensive drive shrinking significantly over time.

Re:wake me up when

[mysidia]

A spinning disk cannot do an infinite number of read-write ops. There is some (very high) number of read-write ops it can do before the mechanical components, read-write heads will deteoriate or fail.

There is also some limit to the amount of time that spinning disk can be powered and spinning, before random failures can occur.

Also, that number is not precisely predictable, nor is it the same or necessarily similar for every hard drive. Sometimes mechanical failure of a spinning disk happens less than a year after its purchase.

Other disks last 10 years of the very same model.

At least with SSD failure that occur are probably not mechanical: they are more likely to do with a manufacturing defect in the electronics on the board, gamma radiation, EMP, or electrostatic discharge, than some random unpredictable event.

Re:In 5 years

[Lonewolf666]

The other respondent who said you should have blamed the managers was more reasonable. Unfortunately current management theory claims that managers don't need to know anything about what they're managing. So the individual managers, themselves, probably aren't to blame.

They were probably told at some point that the current solution would not last beyond the year 2000. That is where even under the current management theory, ignoring the problem becomes their mistake.

In practice, however, I have seen it happen too. Some kind of short term pressure leads to a fishy compromise, and later on management refuses to be bothered by engineers who want to revisit the issue. Until a major breakdown happens or is at least imminent. At that point, lots of money and work goes into fixing things quickly...

Re:Tiered Storage - Software joining SSD+HDs?

[thms]

ZFS offers this already, they call it the L2ARC, you can read about it here: http://blogs.sun.com/brendan/entry/test

Oh, wow! I am even more impressed by ZFS now. Thanks for sharing, This is probably worth a separate Slashdot story.

Re:In 5 years

[scot4875]

You are aware that two bytes (16 bits) is capable of storing 65536 distinct values, yes?

Do you really think that's enough resolution to store any arbitrary date or time? 8 bits isn't even enough to count the number of minutes in a day or days in a year.

And your second paragraph is irrelevant. If disk storage is expensive enough (which it was, back in the day), it doesn't matter what percentage of the drive is used to store the data. If you can prune the data down to save several MB worth of storage, you do it. If data is expensive to transmit, you prune it down to save on transmission costs. If removing/inferring 10MB worth of data will save your organization $10,000 worth of hardware or bandwidth costs, you do it.

--Jeremy

Re:In 5 years

[jon3k]

We had a manager that tried moving to using disk based backup. If you have to rotate the disks, stop now, don't even try. We lost more backplanes and disks in the first 6 months than you would believe. Either put the disks somewhere else and replicate the data to them or use tape. Hard drives are not meant to be moved around constantly.

Re:In 5 years

[jon3k]

Unix Timestamp

Re:In 5 years

[fullfactorial]

What's so forgiving about the magic white smoke getting out of a hard drive after a head crash?

Mechanical failures can be predicted by SMART, and even prevented (e.g. sudden motion sensors that park the heads).

If you DO have a head crash, data recovery services can take the platters out and recover any data not directly affected by the head crash. Some companies offer NAND data recovery, but the current state-of-the-art appears to be more complex, expensive, and uncertain than conventional hard drives.

Re:In 5 years

[fullfactorial]

I was under the impression that SSD's remain readable forever even after they become unwritable.

I have heard the same regarding SLC, but I'm not sure if it applies to the cheaper MLC drives discussed in TFA. Any experts care to weigh in?

Re:In 5 years

[Fjan11]

You could even follow the Unix model and do pretty well keeping the date and time (down to the second) in a single byte...

You are confusing bytes (8 bits) with words (32 bit or 64 bit depending on the system). A byte will only hold 256 discrete values so you could perhaps store the year in there but certainly not an entire date, let alone the time as well.

There was a time once when converting an integer value to and from ascii was prohibitively expensive (hard to imagine now), so it was not unheard of to store a year as 4 bytes.

Re:In 5 years

[gangien]

don't think it's possible? challenge your politician, and see how many laws they can write to protect us from exploding harddrives, or drives that can break, or get stolen or.. who knows.

Re:In 5 years

[Barny]

You can go a step further, using a BartPE you can map folders like "UserData" and "Users" directly to a HDD, throw in "Program Files (x86)" and you can fit windows down into a 32G SSD, create a 3rd program files folder "Program Files (x86SSD)" and dump any small programs in there you want to run really fast :)

As to backing up over USB, eek, its cheap and all, but eSATA is a much nicer way to do it.

Re:In 5 years

[Shark]

The most elegant solution to the energy crisis I've seen in quite a while...

SSDs fail on write?

[Chuck Chunder]

It's my understanding that SSDs tend to fail on write which is detectable allowing the SSD to both be aware of the problem and write the data to another cell. As such I don't think failures should be either less predictable or more unforgiving than with mechanical disks.

Re:SSDs fail on write?

[bill_mcgonigle]

It's my understanding that SSDs tend to fail on write which is detectable allowing the SSD to both be aware of the problem and write the data to another cell. As such I don't think failures should be either less predictable or more unforgiving than with mechanical disks.

The failure scenario I've seen with flash involves rebooting the machine and finding the device to have bad blocks. :(

Re:In 5 years

[jon3k]

And hard drives have "mean time to failure" ratings. That doesn't stop us from selling (or buying) them. You clearly have no concept of the actual limitation of writes on SSDs.

Re:Tiered Storage - Software joining SSD+HDs?

[ImprovOmega]

EMC SAN arrays already do this. It's probably only a matter of time before you see something similar on performance desktops.

Re:In 5 years

[mirix]

I've had a few linux and bsd boxen run off of CF cards, but even the fast ones are so bloody slow compared to even a slow harddrive. I suppose the lack of a buffer might be part of the problem - or do CF cards have small buffers of SRAM?

iowait drives me crazy.

Re:In 5 years

[Mike Buddha]

Pity the lesson of Y2K went unheeded - where every COBOL programmer was paid whatever they asked to fix their code, but after should have all been taken out to a field and shot in the head.

Why shoot the programmers? Why not shoot the managers too ignorant to modernize their code base?

It's not ignorance, or stupidity that's at the root of this. It's economics. Cost vs. benefit analysis. Q: Why replace a system that's been working well for 30 years? Because you want to satisfy your techie ego with the hottest newness available? If the system is failing, then by all means replace it, but if ain't broke, fixing it is entirely a waste of money/time. Mainframes are exeedingly fast in high transaction environments, even by today's standards.

Re:In 5 years

[Mike Buddha]

Yeah, they're a great way to back stuff up... until you have to actually retrieve the data. Then you discover which of your tapes are unreadable, which ones are for the tape drive that doesn't work any more, which ones are scrambled beyond any hope of utility. Get a RAID Array. Do your backups every day and check the integrity of the entire setup every time. Much faster and more reliable. You don't have to wait until D-Day to find out your backups are hosed.

Re:In 5 years

[Sabriel]

For us kids, can you explain how to keep the "date and time down to the second" in a single byte each? Because unless you're using a byte larger than 8 binary bits, for a total of 256 possible values, I'm interested in how you're fitting even a year's worth of days (365), let alone a day's worth of seconds (86400) in that space, nevermind multiple years... what am I missing?

Re:In 5 years

[mirix]

How do you plan on sticking a whole date in 8 bits?

Re:Tiered Storage - Software joining SSD+HDs?

[bill_mcgonigle]

Oh, wow! I am even more impressed by ZFS now.

It is mighty.

This is probably worth a separate Slashdot story.

Tuesday Jul 22, 2008

Re:In 5 years

[AliasTheRoot]

You are right, eSATA is better - it's what we are using. The cradle is an eSATA/USB hybrid & the controller is a LaCie eSATA PCI card (only one we could find at reasonable prices with win2k3 drivers)

There's not much need for using something like BartPE on Vista/7 really - you just right click drag folders from drive to drive.

Re:In 5 years

[evilviper]

You are confusing bytes (8 bits) with words (32 bit or 64 bit depending on the system).

Yes indeed, I was. A single byte would, however, be practically sufficient for a non-astronomical range of years. And I still maintain the parent's assertion of one byte per decimal digit is insane, and yes,

it was not unheard of to store a year as 4 bytes.

Perhaps, but likely not if you are as space constrained and overloaded as the given scenario.

Re:In 5 years

[DJRumpy]

OS's like Linux typically don't use a whole lot of swap given the typical amount of ram in use these days. Also, video editing tends to hit a bottleneck on the CPU, not the disk subsystem. I would think situations that need to read large amounts of data, or a large number of smaller files would benefit most. Things like loading game levels with large 3-D textures, opening large files in Adobe, database maintenance, system boots, etc.

Re:In 5 years

[selven]

Unix time going up to 2038 is 2^31 seconds, or 4 bytes.

Innovator's Dilemma

[rschuetzler]

This is the Innovator's Dilemma. Spinning mechanical hard disks will continue to get larger and larger, but they will be far above the requirements of most users. Some people will want a 4 TB hard drive for their laptop, but for most people 500 GB is more than enough. SSDs will get larger and cheaper until they fit the average consumer's taste. Also, SSDs are faster and more energy efficient, so they will be providing additional benefits on top of size. Thus, my bet is that SSDs will probably move to replace spinning hard disks over the next 5-10 years.

Re:In 5 years

[Anpheus]

Decreasing clock speed with a CPU usually allows a pretty good decrease in voltage, and as you said, the power consumption decreases with the square of voltage.

So if you decrease the clock speed by 10% and that lets you drop the voltage by 10%, you get a ~30% drop in power consumption.

Of course, if you have a fixed voltage or a fixed frequency the equation comes out differently, but anyone who is seriously underclocking for power usage and ignoring voltage is, well, not seeing the forest for the trees.

This is what lead Anandtech to confusingly claim that clock frequency cubed relates to power consumption. In practice it can be close to that, but it is as you said, more complex.

Re:In 5 years

[mrsteveman1]

In standard ATX cases.

I'm well aware that they can run for longer than 3 years, telling me about your specific drives doesn't mean much. Drive quality is highly variable, some of them are good, some of them aren't. The ones I've purchased, which were always at the sweet spot of price/GB at the time, have all stopped working in one way or another within a fairly short period of time.

The post i responded to was a crack about SSDs failing after 5 years because they ran out of write cycles. I've seen them do exactly what i mentioned BEFORE the three year point, and i don't trust them beyond that point anymore.

Re:In 5 years

[mrsteveman1]

What the hell are you talking about?

Laptops and dual bays

[FishTankX]

I really wish that laptops would come with one 2.5" bay and one 1.8" bay.

Right now i'd love to throw an SSD into my laptop for the OS, but am highly portable with my laptop (it travels roughly 6 or 7 miles a day, often in 3 or 4 legs) and don't want to deal with a cramped internal storage situation. If I could simply get a 500GB internal spinner and an 80GB internal flash drive, I would be in heaven. Note to laptop makers, please make this happen. You could even just throw it into the monitor enclosure. But i'd love an extra 1.8" slot.

Re:In 5 years

[mrsteveman1]

Bah, i've seen spinning drives do what i mentioned before, not SSDs :)

Actually, the title was CmdrTaco's not mine

[robertgiam]

I wrote the original story. In the article, I didn't talk about optical storage. Slashdot chose to use a different title over here. But, now that I think about it, I'm not so sure that BluRay will be as long lived as DVD. But it won't be because of SSD. It'll get knocked off by streaming and download services. But pricing is a major issue there. If studios allowed for better rental terms for 1080p VOD from Amazon, Netflix, etc. BluRay would be suffering now - at least amongst folks with 10+ Mbps Internet connections. You can already get high quality downloads to your DVR, PS3, or other box. So better rental terms (like 72 hours instead of 24) + More pervasive high speed broadband + BluRay/DVD-like Interactivity (languages, subtitles, commentaries, etc) = no need for BluRay.

Not going to buy one anyway.

[Hurricane78]

As long as they still wear out that fast, I’m not putting any data that’s worth anything on one of them. No thanks.
Also, it’n not enough that they get on the same price/performance level, to justify a change. They have to be better.
And that means at least 2GB SSD disks for less than 100€, with the same or better reliability and performance.

Re:Not going to buy one anyway.

[Hurricane78]

Ok, stupid typo. I think it’s obvious that I meant 2 TB.
Man, I still remember buying an actual 2 GB one for that price... :/

Re:In 5 years

[theskipper]

Here's an alternate article from Ars. A little less technical but explains things clearly:

http://arstechnica.com/microsoft/news/2010/03/why-new-hard-disks-might-not-be-much-fun-for-xp-users.ars

Some

[zogger]

I can think of some I watched back then on the 50s, for loose, very loose, definitions of science fiction, although the last two in the list were pretty good, more..hmmm..science fantasy/horror.

Superman
Flash Gordon (One of my cats is named after Ming...the Merciless, Flash Gordon's nemesis. He's a rather aggressive and "catcho" tomcat)
Rocketman

One Step Beyond
Kraft Mystery Theater

There's probably more, and certainly a ton of sci fi movies from back then, but those are the TV shows I remember readily.

Re:Some

[mcgrew]

I diddn't consider Superman as "science fiction", even when I was a kid I considered it "juvenile fantasy" (Superman and other comics helped me learn to read). Flash Gordon movies may well have been on TV by the fifties and probably were in some markets, but I was referring to TV shows in particular.

One Step Beyond, Kraft Mystery Theater, and Twilight Zone often had episodes I'd consider science fiction, but for the most part they didn't have what I considered science fiction.

Of course, I probably have a stricter view of what is science fiction; I never considered horror movies to be sci-fi, although most people do.

I do remember (vaugely) one sci-fi show that used puppets. There was Dr. Who, but we didn't get that in the US until way later.

Re:In 5 years

[Barny]

Except when they are folders that cannot be moved/deleted while the user is logged in, such as Program Files, UserData and Users ;)

Yeah, the eSATA/USB combos give the best of both worlds, speed and in a pinch can be plugged in via usb if eSATA isn't there. Check out eSATA controllers based on the SilImage line of chips, they provide chip level driver support for all windows server OS and are reasonably fast/cheap.

Re:In 5 years

[owlstead]

That was the shortest and clearest reply possible to that stupid point. Well done!

Re:In 5 years

[xQx]

The flaw in this logic is exemplified by the fact that VCR's still cost more than DVD players.

HDD's are expensive to create because they are technical and complicated. Conversely, The pharmaceutical business model applies to SSD - the second pill (/chip) costs 30c.

The saving grace may be that at least HDD's are reliable. (Certainly compared to optical media) .. SSD is unproven.

Re:In 5 years

[jbengt]

Why are you assuming that you would be using the same resistance for different amps? Instead, you design in larger wires for more amps and better insulation for higher volts.

My irrelevant thoughts

[Eil]

From TFA:

It tuns out if you look at Intel's and OCZ's new offerings, they're inexpensive simply because they chintz out on storage capacity. Intel's $125 SSD stores 40GB. OCZ's sub-$100 SSD is 32GB. So the cost per gigabyte hasn't really gone down - it's still about $3/GB. At that price, that sub-$200 1 Terabyte Western Digital hard drive would cost you $3,000 using similar SSD technology.

This has always been the argument against SSDs, and it's always been wrong. Pundits are under the impression that it should be possible to get speed, capacity, and affordability all in one go. To use a car analogy, they're asking for a 12-cylinder sports car that gets 40mpg and costs under $30k.

For as long as there have been SSDs, consumers have been waiting for lower-capacity versions that were affordable. It's not that hard to do: just take your "low-end" 160GB version, leave a few chips off the PCB, and voila. The manufacturers so far have been hitting the overclocker and enthusiast crowd who will pay any amount of money for the latest and greatest and the companies are just now realizing that hey, average Joes might buy these things too if we can meet a reasonable price point.

I've personally been waiting for an affordable SSD for my laptop and desktop machines but so far the options have been:

1) A mini-PCIe thing that barely holds an OS and doesn't perform any better than a mechanical disk
2) A fast 2.5" SATA SSD with about 4x more space than I need and costs 2x more than I'm willing to pay

I'm one of those people who doesn't store a crap-ton of data on my computers. 95% of the data on my computers is OS and applications. All of my important or bulk data goes on a file server which is accessed over the network and hence does not need to be available within less than a milisecond.

The day someone sells a fast, reliable, low-capacity SSD (20GB is fine) for under $100 is the day I'll buy three.

Re:In 5 years

[mattack2]

AFAIK, CF cards are basically IDE.. that's why the Apple II Card does both CF & IDE.. http://dreher.net/?s=projects/CFforAppleII&c=projects/CFforAppleII/main.php

So if you already have an IDE controller, I am under the impression that the 'adapter' is fairly simple.

Re:In 5 years

[maino82]

The problem is that you can't always tell when something is failing. A couple of examples from my own personal experience on both the software and hardware side of things...

My bosses have been of the mind for years that you use something til it breaks then you replace it. I've insisted that we take a more proactive stance and regularly replace things, not only to make sure that we have workable hardware but also to make sure we're taking full advantage of all tax breaks available (different things depreciate differently, and after certain periods of accelerated depreciation it doesn't make sense to hold onto old hardware). Long story short, I initially lost that argument and our server died, leaving us out of business for a week. Did we still have things we could do? Sure, but since our business revolves around AutoCAD and producing working drawings on a schedule developed by the architects we work for, this put is in a position to not be able to produce anything that we could actually get paid for. 10 employees at an average of $1000 per day wasted time and our company was essentially out $50,000 of productive work because we didn't spend $3-4k on new servers when we should have. Now we have a firm IT equipment replacement policy in place as a result, but we had to learn the hard way that "use it til it breaks" is not the best way to go and, in the long run, just doesn't make economic sense.

My second anecdote involves AutoCAD again, but in this case we had updated to the latest version of AutoCAD, but we were still using tools, blocks and details developed with a version that was 5 years old because, well, they still worked, right? In the mean time, AutoCAD had developed a wide array of tools and features designed to vastly increase productivity that none of our tools took advantage of because a) no one had kept up on AutoCAD enough to learn all the tricks and b) because we didn't want to invest the time to develop new tools. When a summer intern came in and showed us how cumbersome our old tools were compared to how streamlined and convenient they could be we immediately began updating our libraries. Was it vanity on our part or a need to have the "hottest newness available?" No, it was because we wanted to catch up on all the features and time saving things we'd been missing out on. Is this going to be the same for every piece of software out there? Probably not, but just because you've been doing it the same way for years in no way means you have to continue doing it that way when it may be much more efficient to use the new hotness.

I do agree that updating for updating's sake is not always the best course of action, but sitting on your laurels doing things the same way you've always done them because that's the way they've always been done is also not a very good idea.

Re:In 5 years

[StillAnonymous]

20 years? After 20 years, you're likely to have problems finding a technology that still exists to read that medium. Think about what you were using 20 years ago today. Floppy disks most likely.

What you need to do is migrate your backups to newer media technologies every few years. This way you'll never need to worry about the media itself having to last 20 years, plus you'll get the advantage of it taking up less physical space (a backup that took 100 floppies in 1995 would easily fit on a small USB keychain now, and would take only a few seconds to transfer)

Best of both?

[shish]

Are there any laptops out there with two hard drive slots, so that I could have the system data and random files on SSD, and videos / VM images / other huge data on the platters?

Re:In 5 years

[slimjim8094]

As I understand it you can't even address particular sectors with a guarantee on a SSD so I wouldn't trust a utility like 'shred' because the wear leveling will have it writing 0's all over the disk instead of over the file. But they're easier to dissolve in acid.

Soooo.... Tie?

Not price drops but spinning media itself

[CAIMLAS]

It's not the SSD price drops so much as it is the spinning media - includingDVD, tape, and hard drives - themselves which are pushing people to SSDs.

Why? The older technologies are a pain in the ass:
1) They don't degrade consistently.
2) They offer poor performance.
3) They aren't getting any faster (a marginal concern in most scenarios due to insanely low RAM prices).
4) Most significantly, their reliability has been horrible for the past year+: everyone's seeing massive failure rates for every vendor in the larger capacity disks, with anywhere from 2% to 50% failure rates (per batch) in early-life.

The only thing the older magnetic/optical storage has to offer is capacity, and as that benefit disappears - due to increasingly large SSDs, lower prices, and unreliable large-capacity hard drives - people will stop buying them.

Of course, manufacturers are trying to push people towards SSDs, too: they're much higher profit margin. Though, just like with the CRT/LCD transition, both the new and old transitional technologies suffer in quality until the transition is complete.

Re:In 5 years

[CAIMLAS]

And certainly cleaner than burning hippies and baby seals!

Re:In 5 years

[CAIMLAS]

Meh, 6 months?

Either you've lived in the same area for some time and have a careless carrier, you play hackie sack with your disks before installing them, or you're misdiagnosing common drive thermal adjustments for failure (incorrectly).

A disk isn't "defective" unless it starts having errors. Are you familiar with SMART? It's not fool-proof but it's certainly useful: if SMART starts to bitch, you've typically then got a problem, but short of a "drive doesn't exist" issue, there's rarely a scenario where a disk doesn't show some sign of digital failure in the event of problems (This has been the case for some time.)

As for the longevity of disks... I've got a stack of 5 20G Seagate drives (same model/capacity) next to me right now. They all work still, and were simply the most common capacity (so that I could mess with zfs on them w/o complaints). I'm still using 100Gb and 120GB disks which are over 5 years old (one is powered on and spun down in my 'always on' system and has a SMART on-time life of over 5 years). I've got servers at work which have disks in that ballpark as well.

Either you've got some serious personal issues relating to disks dying, or you need to go out and read a study or two (Google's should do) on disk longevity. Your personal experience is way out of everyone else's experience.

Re:In 5 years

[gangien]

you know why? because we never realize all the benefits of free markets. So, here's a benefit, that you never see or hear about, but it's there. This would not happen if we had similiar laws and regulations for harddrives.

I need SSDs for Porn!

[jonaskoelker]

I mean really, who needs an expensive big SSD for your porn collection?

Me. Well, kind of. I'd use it in a hybrid way.

For every video and audio file, I'd store the first X miliseconds worth of data on the SSD and the rest on the HDD, where X is chosen such that I can fetch another hunk of data from the HDD within those X miliseconds with probability p, say for p = 99%.

In that way, I would have playback start instantly (the promise of SSDs) and I would store all my stuff cheaply (the promise of HDDs). It won't be optimally fast or optimally cheap, but it'd be like a 90/90 going against a 50/100 and a 100/50 (for certain anally extracted values of 90, 50 and 100).

Let's have a closer look

[jonaskoelker]

FTF WP A:

Wear leveling used on flash-based SSDs has security implications. For example, encryption of existing unencrypted data on flash-based SSDs cannot be performed securely due to the fact that wear leveling causes new encrypted drive sectors to be written to a physical location different from their original location

Fail. Really, this is about deletion, not encryption: you still get to have E(x), but you don't have it rather than x, you get to have both.

And the same issue presents itself on HDDs if you use journaling file systems. And if you don't, you get slow fsck operations, plus your data could be better protected at non-noticeable cost. Your disk may work twice as hard (doubly writing everything), but it does so when you look away and don't notice it.

Re:In 5 years

[FuckingNickName]

If your entire date field takes more than 1 byte, you're doing something wrong.

[...]

--

Yes, I am God. [slashdot.org]

Ah, Young Earth Creationists today...

Re:In 5 years

[xtracto]

huhuhu this made me recall a [bad]... someone told me "it is the question that every engineer must know th answer to", back when I was starting my soft. eng. bachelor degree:

What was first, the masturbation of Men or the masturbation of Women? ....
A: The masturbation of men was first because it is "manual" (hand-operated) and Women's masturbation is "digital".
har har...

Re:In 5 years

[LordLimecat]

How much swap space are we talking exactly?

Re:Tiered Storage - Software joining SSD+HDs?

[LordLimecat]

AFAIK CPU cache isnt used for data, and I dont think youd want to do that regardless.

Re:In 5 years

[aix tom]

The last one I *really* needed was a 2GB disk when the RAM was 256MB

Right now I use a 100GB Disk with 20GB for swap with 8GB physical RAM . and the rest for some other "Temporary" Data.

In 20 years

[stonewolf]

No one will remember drives at all. They will be just another part of the chipset like the sound card and Ethernet.

What I am really curious about is whether it will be drives or video cards that disappear first. Storage technology is getting cheaper faster and only needs to optimize price per bit. While video cards are still not capable of all the things we want them to do and have to optimize many different factors. So, I'm betting on storage disappearing first, but I could be wrong.

Stonewolf

Re:In 5 years

[toddestan]

You have to remember that a write on a SSD is actually a two step process. First, the block must be erased, then the new data can be written to the block. If the failure occurs on the first step, then you could still have your old data. But if it occurs on the second step then you have some useless blank areas that can't be written to. The problem I see is in almost all cases you're writing a whole lot of blocks, and chances are most of them would still be working with only a few that failed, leaving you with an incomplete, corrupted file unless you verify all your writes.

This is all assuming the media is what gives up the ghost. I haven't had an SSD fail on me yet, but all the failures of my USB flash drives seem to be with the controller/interface chip and not the flash itself, leaving a completely dead drive with little hope of recovery (at least cheaply).

Re:Tiered Storage - Software joining SSD+HDs?

[Just Some Guy]

Tiered Storage - "on the fly" support for that is something I would love to see:

Great idea! I'm using that today on FreeBSD with great success. Furthermore, the idea works well enough in practice with real machines that even a cheap USB flash drive gives a nice boost.

Re:In 5 years

[bingoUV]

People recommended tricks like XORING two pointers together to save space

While I see your point, any example where XOR of two pointers is of any practical utility?
thanks

Re:In 5 years

[pnutjam]

I've been rotating an archive drive for 2 years, we had a set of 4 and yes, one of the cadi's broke a tab so it wouldn't lock into the server, but it still worked fine, as did the server side of the hardware. After 2 years we upgraded the server and the archive drive size. But the old archive drives are shelved and ready, they still work. New server has been in use for a year.

Yeah, if I want to store something in a vault for 50 years, tape is probably better. If you want to store company data, probably going back 2-5 years, disk is way better. It's lightning fast (compared to tape), to transfer that data to a new drive if you necessary. The number of times you can write to a disk is way higher then the number of writes any tape supports.

I don't see the benefit of tape for most companies where data is a living thing.

We us Unitrends, check it out.

Re:In 5 years

[pnutjam]

to be fair, it would take afew seconds to load each floppy, so 100 floppies, times 10 seconds (loading and copying)= 1000 seconds / 60 = less then 17 minutes to move all your data to a new media, and next time you need to do so it would be more like 3 minutes.

Re:In 5 years

[petermgreen]

On the other hand, the current drain from these things is probably so low that I'm not sure if it will make enough difference to worry about. :-)
Agreed, reading the datasheet for the X25-M the typical power consumption its given as 75mW idle and 150mW active. That is just a fart in the wind by PC standards.

Re:In 5 years

[bluefoxlucid]

And yet, they run high-voltage low-amperage power in the EU and over HVDC lines to prevent wires from melting from huge current...

Re:In 5 years

[Pentium100]

Because we are talking about different things.

Transmission losses are P=I^2*R, so if you halve the current you have 4 times less power heating the wire.

The voltage on the wire is quite small (U=I*R), even though the voltage between the wire and another wire (or ground) is 750kV or more.

However, as to the actual device that uses power, it can use whatever voltage and current it is designed to use.

Now, power loss in a chip depends on the voltage (and frequency, and duty cycle), that's why new CPUs use less than 1V. Flash chips work on 5V or 3.3V and making ones that work on 12V would probably be expensive or impossible, which means that if you want an SSD that works on 12V, the SSD will have a voltage regulator which converts the 12V to 5V or 3.3V.

Powerful DC-DC converters are more efficient than small ones, also, dual conversion (220V->12V->5V) is less efficient than converting straight from 220V to 5V.

However, SSDs do not use a lot of power, I can't say how much exactly, but my 750GB 7200RPM hard drive uses ~15W, so a SSD would use less than that, but let's say 15W.

At this power, it's 3A at 5V or 1.25A at 12V. The wires from the PSU to the SSD are short and quite thick so they won't lose much power as heat. A SSD that works on 12V (and has a DC-DC converter to step it down to 5V) would be less efficient because you would lose more energy in heating the DC-DC converter than you would lose passing those 3A on the wire.

Re:In 5 years

[bluefoxlucid]

Transmission losses are P=I^2*R, so if you halve the current you have 4 times less power heating the wire.

The wire is a resistor...

I didn't think of how to convert 12V to 3.3V or less at the other end, I just don't want high amperage running through the power supply and making it hot.

Re:In 5 years

[HiThere]

One place it was used was in doubly linked lists. If you XOR the next and prior pointers together, then you only need to save one word of pointer information..at the cost of more complex retrieval and updating algorithms.

I think I got that trick from Knuth, but I'm not certain.

Re:In 5 years

[HiThere]

I know what you mean, but yes, 20 years.

20 years isn't long enough, but it's a good start. In the early days CDs promised a lifetime in the centuries, but that was when data was written on holes burned in metal foil (gold?) that was sandwiched between two glass disks.

P.S.: I've had mag tape that held up for 20 years. True, it was only at 800 BPI odd parity, and there were bad spots. But *almost* all the data was recovered. The even parity tapes were a bit less forgiving. (In 1980 something we needed to compare the current data against 1960 census data.)

Re:In 5 years

[Pentium100]

So you'll make the DC-DC converter on the SSD hot instead? Also, you will lose more energy heating both the SSD and the power supply, even though the power supply will be a half degree colder.

Re:In 5 years

[Just Some Guy]

4) Use for swap space 2-3 Years

Ummm, that's actually what you want to use your newest drive for. Newer drives are almost always faster, especially compared to ones 4-6 years old (as the swap drives would be in your case). In the event that you're actually paging out, wouldn't you rather it be to a partition on a lower-latency drive with better throughput?

Re:In 5 years

[godefroi]

So the performance-critical always-accessed stuff on cheap slow storage, and the rarely-used stuff on fast expensive storage.

I like your style.

Re:In 5 years

[godefroi]

OS's like Linux typically don't use a whole lot of swap given the typical amount of ram in use these days.

Neither do OSs like Windows.

video editing tends to hit a bottleneck on the CPU, not the disk subsystem.

Admittedly I've only ever done a small bit of video editing (home movies from a DV camera), but every single bit of it was limited by IO. Never, ever, by CPU, and that was back in the days where Athlons and P4s came in sub-2gHz speeds. If you'd have said "video rendering" then I'd have believed you.

Re:In 5 years

[petermgreen]

proper pressed CDs do have a very long life. Unfortunately CDRs can't match it since they use dye layers instead of pits.

Standalone vs. Add-in

[itomato]

The figure I quoted was from a BTO session at HP.

This is a bare internal drive, and actually I may be $500 off or thereabouts.

Even more to the point, if the drive is $1,300..