The Limits To Perpendicular Recording

peterkern writes "Samsung has a new hard drive and says it can now store 667 GB on one disk, which comes out to be about 739 Gb/sq. in. That is more than five times the density when perpendicular recording was introduced back in 2006, and it is getting close to the generally expected soft limit of 1 Tb/sq. in. It's great that we can now store 2 TB on one hard drive and that 3-TB hard drives are already feasible. But how far can it go? It appears that the hard drive industry may start talking about heat-assisted magnetic recording again, soon."

Comments

Heat-assisted magnetic recording?

[FooAtWFU]

When the only tool you have is a HAMR, everything looks like a nail.

Re:Heat-assisted magnetic recording?

[Anonymous Coward]

When the only tool you have is a HAMR, everything looks like a nail.

I nailed your mom last night, perpendicularly.

Re:Heat-assisted magnetic recording?

[QuantumRiff]

No..
When the only tool you have is a Hammer, every problem looks like a hell of a lot of fun....

Re:Heat-assisted magnetic recording?

[noidentity]

When the only tool you have is a HAMR, everything looks like rust.

There, fixed that for you.

Re:Heat-assisted magnetic recording?

[cgenman]

My laptop's hard drive already utilizes heat-assisted magnetic erasing, though it tends to work on an entire drive at a time.

Re:Heat-assisted magnetic recording?

[Anonymous Coward]

The HAMR is my penis.

Re:Heat-assisted magnetic recording?

[Local ID10T]

oh no. I can picture the marketing now...

HAMR time!

and now I can't get it out of my head...

Re:Heat-assisted magnetic recording?

[Anonymous Coward]

So would a hammer.

Re:Heat-assisted magnetic recording?

[smallfries]

Don't you just love Apple innovation...

Re:Heat-assisted magnetic recording?

[Yvan256]

The HAMR is now diamonds.

Re:Heat-assisted magnetic recording?

[jd]

Let one who has understanding reckon the number of the Sony, for it is a gigabyte number, its number is six hundred and sixty seven.

Re:Heat-assisted magnetic recording?

[Anonymous Coward]

When the only tool you have is a hammer, every problem looks like a thumb.

Re:Heat-assisted magnetic recording?

[Maarx]

Re:Heat-assisted magnetic recording?

[Just Some Guy]

I'd HAMR in the morning. I'd HAMR in the evening. I'd HAMR all over this land. And that works out to about 1.12*10^19 GB for the surface area of the US.

Re:Heat-assisted magnetic recording?

[Anonymous Coward]

Give him some slack He downloaded some porn and the EFI instructed the drive to HELP him delete such nasty and unholy content from his white and pure piece of hardware.

Get Perpendicular

[Jazz-Masta]

A simple video to explain perpendicular recording!

http://www.youtube.com/watch?v=xb_PyKuI7II

Re:Get Perpendicular

[MikeMayer]

Love it.

Re:Get Perpendicular

[Anonymous Coward]

Clicky link for the lazy.

Re:Get Perpendicular

[ShakaUVM]

"I'm dancing! I'm dancing!"

I wonder if the guys who made Borderlands had seen this video.

TFA is unreadable.

[FooAtWFU]

However, more density also provides a way to higher capacity 3.5" drives, which means that Samsung is now able to build 2.7 GB and 3.3 GB hard drives with four or five disks, respectively. Such drives are rather unlikely however, as we would expect the density to grow to 750 GB per disk, which could enable 4-disk 3 GB drives.

Oh, wow, a 3-gigabyte drive! How futuristic!

Seriously, what sort of monkey messed the article up this badly?

Re:TFA is unreadable.

[Jazz-Masta]

However, more density also provides a way to higher capacity 3.5" drives, which means that Samsung is now able to build 2.7 GB and 3.3 GB hard drives with four or five disks, respectively. Such drives are rather unlikely however, as we would expect the density to grow to 750 GB per disk, which could enable 4-disk 3 GB drives.

Oh, wow, a 3-gigabyte drive! How futuristic!

Seriously, what sort of monkey messed the article up this badly?

This is slashdot, in the 12 years I've been wasting time here, I am more surprised when they get a story with all of the facts, spelling and concepts correct!

Re:TFA is unreadable.

[Monkeedude1212]

This is the internet. Facts, spelling, and concepts are all optional.

Re:TFA is unreadable.

[Anonymous Coward]

To be fair, not many people are used to writing GB yet. After a decade of writing GB getting them to start using the abbreviation GB will take time.

Re:TFA is unreadable.

[Anonymous Coward]

kdawson posted TFA. This explains everything.

Re:TFA is unreadable.

[Culture20]

Oh, wow, a 3-gigabyte drive! How futuristic!
Seriously, what sort of monkey messed the article up this badly?

Yeah, seriously. I thought we agreed on using Gibibytes from now on, right?
I actually make the TB/GB mistake regularly, and people either don't catch it, inserting TB because they know I'm talking about a low (<10) number or they just don't bother to mention it because they can reason out what I meant. That's one of the nice things about having 1000 or 1024 as a separation point in the naming hierarchy.

Re:TFA is unreadable.

[Jah-Wren Ryel]

To be fair, not many people are used to writing GB yet. After a decade of writing GB getting them to start using the abbreviation GB will take time.

I see what you did there,

Re:TFA is unreadable.

[dclozier]

So what your saying is that you've never been surprised. :D

Re:TFA is unreadable.

[lethalwp]

haven't you read the EULA?

Re:TFA is unreadable.

[Yvan256]

Georgibi Bush?

Re:TFA is unreadable.

[CastrTroy]

Also, they all assume we need more space. Most of us don't. Especially not consumers. Maybe in industry they need more space, but there's a solution for that. More drives. There hasn't been anything that's come around in the last 15 years that has required more storage than video. So, the most space hungry thing we have is video. Everything else on my personal computer pales in comparison. So, Most of the Hi Def Rips I see are around 5GB. Now, if you had a 3TB drive, that gives you 600 Movies. You really don't need to store 600 movies all at once. And with the way internet is going, you really shouldn't need to store any of it, because you can just download it as you watch it. If we all had 50 MBit internet connections (there will be one day), there will be no need to store anything this large on your own personal computer. NetFlix already offers a huge part of their library online for $10 a month. No reason to buy a movie. Just "rent" it again if you want to watch it. I already don't keep most of the stuff I download. I have the space to keep it, but I say, why bother. I'll just delete it, and if I really want to watch the same movie again in 3 months, I'll download it again. Apple and others could probably do the same with music. Charge you $5 a month to manage all your song purchases. You never have to worry about backing it up. You never have to worry about losing music. They really only have to keep 1 copy of each song, and then a list of the songs you have access to. It might take another 10 years for things to really get going, but it's getting to the point where having a lot of content stored on your own computer just doesn't need to be done.

Re:TFA is unreadable.

[Anonymous Coward]

Says 3TB (terabytes=1000GB) not 3GB.

I knew this was a kdawson post...

[elohel]

Most of the time I never comment on how dumb a synopsis is...but HOLY SHIT. I had to log in and comment to just complain about how terrible this is. NEWS FLASH: Technology has finite limits! In other news, fire is hot and humans eat food. More at 11. "It appears the industry may start talking about heat-assisted magnetic recording again, soon." Thanks for actually saying nothing. Your comments to the article are completely useless. This is one of the reasons why slashdot gets on my nerves, what useless junk.

Re:I knew this was a kdawson post...

[Anonymous Coward]

This is one of the reasons why slashdot gets on my nerves, what useless junk.

Then don't read it.

Re:I knew this was a kdawson post...

[Anonymous Coward]

Dawson is that you? Could you pretty please stop sucking? Or at least take it out of your mouth long enough to read what you write?

Re:I knew this was a kdawson post...

[Anonymous Coward]

Bullshit. He's right. Slashdot is a sack of shit sometimes. And instead of simply not reading it, he's complaining in the hopes they get a small message eventually and maybe strive to be better.

In the long run, people talking about what a declining news site Slashdot is should have some effect.

Kdawson should never have been allowed to speak here. Slashdot was supposed to filter out this type of garbage.

Re:I knew this was a kdawson post...

[Anonymous Coward]

In the long run, people talking about what a declining news site Slashdot is should have some effect.

The best plan so far to bring this place back from the dead!

Re:I knew this was a kdawson post...

[Anonymous Coward]

If you don't think that an article about:
A) A new hard drive featuring the highest storage density ever
B) The physical limits of current hard drive storage potential
C) Potential solutions to those physical limits
is worthy of Slashdot, you don't belong here, so let me be of some assistance: www.digg.com

Re:I knew this was a kdawson post...

[Anonymous Coward]

Dude, why do you choose to suffer so? You and all the other apparent masochists who subject themselves to reading that which they find so offensive.

Re:I knew this was a kdawson post...

[hitmark]

and still you continue to read it.

Re:I knew this was a kdawson post...

[elohel]

Dude, why do you choose to suffer so? You and all the other apparent masochists who subject themselves to reading that which they find so offensive.

I believe this is a good question. I continue to read Slashdot because the obtaining of information (news) and the opinions of the few (kdawson) are not the sole reasons for my use of Slashdot.

I find that there is quite the collective of intelligence here. I almost ALWAYS RTFA but never comment. I enjoy reading what other people have written. I'm fairly young - I don't have an all-encompassing world-view; I lack many experiences that life will undoubtedly heap upon me. News is just that...news. Events. Things happen. What makes that news interesting is CONTEXT, which is what many users here at Slashdot offer me. Comments to articles are much more enlightening and provide a much broader learning experience than the article itself in most cases.

So, to simply answer your question..

I don't suffer. It is better to pass through the sorrow so that we may know the good from the bad.

Stop Making It Bigger. Start Making It Faster!

[MankyD]

Stop making it bigger! Start making it faster!

Re:Stop Making It Bigger. Start Making It Faster!

[geekoid]

That what she said.

only visa versa..

Re:Stop Making It Bigger. Start Making It Faster!

[Monkeedude1212]

That's what the SSD market does.

Re:Stop Making It Bigger. Start Making It Faster!

[Kenoli]

The real performance boost comes from higher capacity drives. If you want fast, just get two.

Oh wait..

Re:Stop Making It Bigger. Start Making It Faster!

[toastar]

Stop making it bigger! Start making it faster!

What about harder and stronger?

Re:Stop Making It Bigger. Start Making It Faster!

[EdZ]

No, make it bigger! Bring back 5.25" form factor drives!

Re:Stop Making It Bigger. Start Making It Faster!

[Anonymous Coward]

Hard drives can already read/write around 50 MB per second, the real problem is programmers who insist on scattering all the data for one program around in 1000 different files. Zip that shit up so it can all be read at once, damn it.

Re:Stop Making It Bigger. Start Making It Faster!

[jtownatpunk.net]

Why not do both in separate product lines? Kinda like what they're already going right this very moment. If I want a lot of stuff in one place, I buy hard drives. If I want a small amount of stuff accessed very quickly, I buy SSDs. One division increasing capacity doesn't stop an entirely different division from increasing performance. And those SSDs are increasing in size pretty quickly. The Vertex 2 Pro is up to 240 gigs for under $700. Wasn't long ago that the tiniest, crappiest-performing SSD cost that much. Now that's the price of the biggest and fastest. In another year, the $/gig ratio will be even better along with performance.

So I think fast storage is coming along just fine and I'm happy to have the slow spinning stuff for my "access occasionally" data like audio, video, backups, etc.

Re:Stop Making It Bigger. Start Making It Faster!

[TeknoHog]

That's what She SaiD.

Re:Stop Making It Bigger. Start Making It Faster!

[tagno25]

That's what the SSD market does.

No, the SSD market makes it faster to read, slower to write, shorter write life, and ~10 times smaller. What we need is slightly faster read and write, longer life, the same capacity.

Re:Stop Making It Bigger. Start Making It Faster!

[petermgreen]

The Vertex 2 Pro is up to 240 gigs for under $700. Wasn't long ago that the tiniest, crappiest-performing SSD cost that much. Now that's the price of the biggest and fastest.
I can't comment on fastest but it's far from the biggest. You can get 512GB and 1TB SSDs (though the 1TB ones are desktop form factor) now but the price is insane.

In another year, the $/gig ratio will be even better along with performance.
I sure hope so

Re:Stop Making It Bigger. Start Making It Faster!

[Anonymous Coward]

They're already hard.

Re:Stop Making It Bigger. Start Making It Faster!

[Anonymous Coward]

Right, because zipping up 1000 files into 1 compressed file, and then decompressing it to find the file the app is looking for makes it more efficient. I see now. Good work!

No, but i sort of agree. Too many files fragmented over the capacity of a hard drive makes things slower. less files would speed things up. but just build a bigger circuit board for higher capacity SSDeez! And make them consumer-affordable already bitchez!

Re:Stop Making It Bigger. Start Making It Faster!

[tknd]

No, keep making it bigger. When we had floppy disks, it easy to fill up your floppy disks with documents. With hard drives, that became much harder so hard drives made document storage nearly infinite. As computers got faster, we started listening to mp3s and taking pictures. But it was still possible to fill up your hard drive with mp3s or pictures. The drives kept getting bigger and now at about 2TB, I'd think it would be pretty hard for most people to fill that up with mp3s or pictures, so now for music, pictures, and documents, hard drives are nearly infinite for most people.

But let's not stop there. It is still easy to fill that 2TB hard drive with video like dvds and bluray. So let's keep going to petabytes and maybe then a single drive ought to be enough for anyone.

Re:Stop Making It Bigger. Start Making It Faster!

[Grishnakh]

Theoretically, for many applications, zipping up the 1000 files into 1 compressed file and decompressing it on-the-fly really is faster, and has been for quite some time. Disk speeds haven't changed that much in the past 10-15 years, but CPUs and memory buses have become far, far faster. Since disk seek time and latency is so long, compared to the amount of work a modern (esp. multicore) CPU can do in that amount of time, it frequently makes more sense to compress data and archive disparate files into single larger ones.

Re:Stop Making It Bigger. Start Making It Faster!

[Monkeedude1212]

I don't think we need that much faster write speeds - once your installation and setup is done, you're dealing with peanuts in size of writes, kilobytes, maybe a few megs.

SSD's do exactly what PC's need - a much faster read, at a much smaller volume, with decent enough capacity. Lifetime is something that they aim to mature with development.

Re:Stop Making It Bigger. Start Making It Faster!

[Jah-Wren Ryel]

No, make it bigger! Bring back 5.25" form factor drives!

If they made them cheap enough, I would buy 6TB quantum bigfoots for archival purposes.

Re:Stop Making It Bigger. Start Making It Faster!

[Anonymous Coward]

You joke, but I would love to have a slow 4200RPM half-height 5.25" drive able to store 10-12 terabytes for mass storage. It would be perfect for home media servers where access time isn't all that important, and it would be more than fast enough to handle HD video.

Re:Stop Making It Bigger. Start Making It Faster!

[nizo]

Forget both those, how about more reliable?

Re:Stop Making It Bigger. Start Making It Faster!

[fishbowl]

There should be factors of human proportions that limit the need for exponentially increasing growth at some point.

Human perception in audio has already been passed both in frequency and dynamic domains. Static images are reaching that threshold, and we do already have lossless encodings that pass it. Motion pictures will be the next threshold, and then I suppose holography. So there goes my argument that we can limit the need for exponential growth, oh well.

I think it's funny that you can probably store all known pre-17th century literature and a decent representation of art, music, and architecture of the whole pre-industrial period on a pocketable medium.

Re:Stop Making It Bigger. Start Making It Faster!

[X0563511]

You say slow, but at that density the throughput will still be quick.

Re:Stop Making It Bigger. Start Making It Faster!

[cynyr]

start ripping bluray movies... 30-40GB per...

Re:Stop Making It Bigger. Start Making It Faster!

[Monkeedude1212]

Thats what Hard drives are for!

Re:Stop Making It Bigger. Start Making It Faster!

[Znork]

With the surface area of a 5.25" drive and today's densities you could probably fit 10-20TB in them. I'd definitely buy several if they had an attractive price of around 30-50% of 3.5" price/GB. Speed isn't much of an issue and if it were I'd go with striped SSD's or simply more RAM anyway, but sheer storage capacity is never enough, and if it ever becomes enough, I can certainly use up even more by expanding redundancy.

Re:Stop Making It Bigger. Start Making It Faster!

[gringer]

That what she said.
only visa versa

Is that when she spends exactly the same amount you saved on your credit card?

Re:Stop Making It Bigger. Start Making It Faster!

[Anonymous Coward]

Don't know what SSDs you're not using, but last time I checked even entry level ones from a couple of years ago write twice as fast as any spinner out there now. The new SSDs now are several times faster than that. (Read and write.)

I won't argue on size, but life is even steven. I ditch all spinners after 3-5 years since they are living on borrowed time after the warrenty runs out, so who cares if the standard drives last longer. Only an idiot would trust any kind of drive for the long term.

My pair of orange OCZ SSDs are going on their 3rd year and are still humming along nicely. They'll get dumped into a car PC build I plan on doing and replaced by the latest and the greatest when I rebuild my main PC.

Re:Stop Making It Bigger. Start Making It Faster!

[bertok]

Theoretically, for many applications, zipping up the 1000 files into 1 compressed file and decompressing it on-the-fly really is faster, and has been for quite some time. Disk speeds haven't changed that much in the past 10-15 years, but CPUs and memory buses have become far, far faster. Since disk seek time and latency is so long, compared to the amount of work a modern (esp. multicore) CPU can do in that amount of time, it frequently makes more sense to compress data and archive disparate files into single larger ones.

You'd be surprised.

I've recently had to optimise a compression step in a large system, and I was appalled at how slow most compression libraries and programs are, especially the ones in common use.

Typical (zip) style compression libraries rarely exceeds 10MB/s compression rates, and 20-30MB/s in decompression. That's substantially slower than what most mechanical hard-drives can do, let alone SSDs. In practice, reading or writing a 'zip' file, which includes all MS Office 2007 formats, XPS, etc... will be CPU limited.

There's all sorts of sillyness: many libraries perform IO operations with tiny buffers (4K or less), perform IO synchronously, and don't take advantage of 64-bit instructions, SSE, or multi-core CPUs. Even if optimisations were used, most compression formats are very heavy on unaligned byte and bit twiddling, which is inefficient on modern CPUs.

Re:Stop Making It Bigger. Start Making It Faster!

[Anonymous Coward]

Photographs will keep getting more pixels because there is always more detail and context that could be visible if you had more of them. Same goes for video.

Re:Stop Making It Bigger. Start Making It Faster!

[Anonymous Coward]

not just 5.25" - but full height as well...
So how many platters could we fit in a full hight drive these days?
15-20? more?
Wikipedia lists a full height 14 platter Seagate from 1998 (elite 47) So I'd be betting on 20+ platters
40-50 surfaces, with around a TB/surface...
Very nice!

Re:Stop Making It Bigger. Start Making It Faster!

[petermgreen]

The biggest problem with HDDs is that while sustained read speeds have been going up a bit and capacities have been going up a lot transactions per second has been sitting pretty still. You can't get away from the fact that it takes time to physically move the heads and then wait an average of half a rotation for the data to be under them.

Increasing the rotation speed works to some extent but there are big heat problems with that.

Re:Stop Making It Bigger. Start Making It Faster!

[davidbofinger]

Stop making it bigger! Start making it faster!

What about harder and stronger?

How about a compromise? We'll make it faster and stronger, I'll throw in better, and it will only cost you six million dollars.

Re:Stop Making It Bigger. Start Making It Faster!

[EvanED]

Static images are reaching that threshold, and we do already have lossless encodings that pass it.

I'm not going to comment on resolution, but in terms of dynamic range, I don't think we're anywhere near limits of human perception. Certainly not in anything that approaches a consumer-level device.

Re:Stop Making It Bigger. Start Making It Faster!

[Drathos]

Sorry, but I can't afford $6,000,000 for a hard drive..

Re:Stop Making It Bigger. Start Making It Faster!

[Pentium100]

Actually, I think they should concentrate on making hard drives even bigger, for example, returning to 5.25" size with modern technologies (perpendicular recording etc). A hard drive with 8-16 5.25" platters would have quite high capacity (probably 10-20TB). Also, linear speed would still be pretty fast (the edge of a 5.25" platter spinning at 5400RPM travels about as fast as the edge of a 3.5" platter spinning at 8000RPM)

Oh, and I would buy a smaller, faster drive (say 146GB-300GB, 15000RPM) for stuff like system files too. I do not need 5ms access time on my movies/flac files etc.

Re:Stop Making It Bigger. Start Making It Faster!

[countertrolling]

...zipping up the 1000 files into 1 compressed file...

Yeah, that way you can corrupt a thousand files for the price of one.

Re:Stop Making It Bigger. Start Making It Faster!

[Unoti]

You can't get away from the fact that it takes time to physically move the heads and then wait an average of half a rotation for the data to be under them.

You can, sometimes... this is one of the central ideas that MapReduce and Hadoop are all about: removing disk seek times from the equation and getting the data streaming non stop. Things get a lot faster when the application is designed start to finish to stream as much as possible.

Re:Stop Making It Bigger. Start Making It Faster!

[Ambiguous Puzuma]

Perhaps, but the compression ratio can easily make up for it.

On my ~5 year old computer, using an old version of cygwin:
Searching 16 GB of text files takes about 10 minutes.
Unzipping 1.6 GB of zip files (containing the same data) and searching them takes about 3.5 minutes.

So as long as the compression ratio exceeds 3.5:1 (and I'm likely to read much more often than write) I'm probably better off decompressing on the fly, even using good old "unzip".

Re:Stop Making It Bigger. Start Making It Faster!

[profplump]

That's a limitation of display technology though, not storage capacity. Even if you wanted, for example, 192-bit color (i.e. 10^57 colors), that's not even an order of magnitude more storage than today's typical image storage system.

Re:Stop Making It Bigger. Start Making It Faster!

[x2A]

"...after the warrenty runs out"

Yes I also feel a lot safer knowing that if my hdd fails destroying all the data on it, at least I can get the drive replaced for free, that why I only ever use drives within their warrenty too. I also completely ignore the bell curve failure of drives that shows greater risk involved with a brand new drive than a drive that has lasted a few years with no signs of problems.

Re:Stop Making It Bigger. Start Making It Faster!

[mcrbids]

5.25 drives, especially full height, outta hold a *lot* more than a mere 6 TB - I'm thinking 20-30 TB.... 5.25 full height is *big*.

Re:Stop Making It Bigger. Start Making It Faster!

[Anonymous Coward]

Quantum Bigfoots were never full height.

Re:Stop Making It Bigger. Start Making It Faster!

[cgenman]

Considering my laptops seem to eat traditional disk-based HDD's in about 2 years, I'd consider SSD's the longer-lived option.

Re:Stop Making It Bigger. Start Making It Faster!

[l0b0]

At a recent CERN colloquium, Evangelos Eleftheriou of IBM Research advocated using Flash cache & SSDs as an extra storage layer between RAM and old-fashioned harddisks. I think the argument was simply that price and performance is between both, and is foreseen to stay that way for the foreseeable future, so it's the most economic option for now.

Re:Stop Making It Bigger. Start Making It Faster!

[Jeprey]

All the women I know say size really does matter. Fortunately I have no problems in that department.

There are limits to speed by the inherent design of HDs though be aware that the raw data rate of most heads today is already in the 2 Gb/sec range. Which if you never seek the head, you'll be a-ok. :-)

Re:Stop Making It Bigger. Start Making It Faster!

[dwinks616]

Or you could buy 3 3TB drives, put them in a raid 5 for your 6TB total and not lose 6TB when 1 drive fails...

Re:Stop Making It Bigger. Start Making It Faster!

[Anonymous Coward]

Don't be daft, punk!

Re:Stop Making It Bigger. Start Making It Faster!

[Coren22]

The problems that killed the 5.25 were power draw and speed. Nowadays, I would expect that the power draw would be lowered due to the use of plastic substrate with a film of metal rather then the old time solid metal platters, as well as the speed isn't such a big deal, just throw a huge buffer on it, or SSD buffer and you are all set. I could definitely deal with a 20TB 5.25, or 40TB Full height 5.25. The hard drives in my home computer already take up that space anyways:

http://www.newegg.com/Product/Product.aspx?Item=N82E16817121405

Heck, they even take up three slots.

I actually still have a Full Height Seagate drive, SCSI, 47GB, it takes about 5 minutes to spin up, but had enormous space for the time period. Unfortunately, a capacitor fell off of it, and I haven't had the motivation to fix it as I have so much storage on all my machines.

Re:Stop Making It Bigger. Start Making It Faster!

[The Wild Norseman]

All the women I know say size really does matter.

So how is your mom, anyway?

Re:Stop Making It Bigger. Start Making It Faster!

[Anonymous Coward]

Our work is never over.

Re:Stop Making It Bigger. Start Making It Faster!

[Anonymous Coward]

If we get to deeper, we've got a problem.

Re:Stop Making It Bigger. Start Making It Faster!

[damnfuct]

Put two sets of heads on the drive. Confine one to an inside annular region, and the other to the outside annular region. Put the actuator arms on opposing sides of the platter to make it easy. Then you'd probably have a seek time similar to a 3.5" drive.

Re:Stop Making It Bigger. Start Making It Faster!

[rrohbeck]

Bigfoot was only economical at the end of the 5.25" era, when the platters could be had for cheap because the demand was going away. Today nobody makes 5.25" platters any more.

SQUID's next?

[Tisha_AH]

There is only so much you can pack into little magnetic domains. It is dependent upon how small of a grain (dust speck) you can individually magnetize, signal/noise ratio to read back that magnetic field and the sensitivity of the pickup head. I can see the day coming when there is a small near-room-temperature superconductor (SQUID) pickup head to do read/write operations. The tradeoff is going to be when you get that small, a single cosmic ray particle can flip a 1 to a 0.

Re:SQUID's next?

[jandrese]

I thought the downside was the need to keep a tank of liquid nitrogen nearby to cool the heads down to superconducting temperatures?

"Near room temperature" superconductors typically still need to be well below zero to operate. Until we get one that can operate at 100C, this idea is a nonstarter for all but maybe some kind of industrial or scientific application.

Re:SQUID's next?

[DarkKnightRadick]

100C is awful hot.

Re:SQUID's next?

[JustinRLynn]

You could always attach a cryocooling apparatus to the head and head area. If it's small and lightweight, it could potentially be miniaturized to fit inside the drive's casing itself. If not, well.. you remember those people who put in massive refrigeration systems in order to over-clock their systems? Well, they're still out there.

Re:SQUID's next?

[petermgreen]

Given that HDDs in underventilated cases sometimes reach 60 or so and you want some safety margin it sounds like you'd want one that would work up to 80 or so.

So while I think 100 is higher than needed it's not completely out of whack.

Re:SQUID's next?

[DarkKnightRadick]

I didn't realize it got that hot in there. I figured they were at most no more than 10 degrees above room.

Re:SQUID's next?

[Grishnakh]

You can use small, localized cooling systems to keep the temperature down, such as Peltier coolers.

Re:SQUID's next?

[X0563511]

I've had a drive burn me when I tried to remove it. Yes, it actually burned me.

I then proceeded to slap the idiot who built that server into the stone age.

Re:SQUID's next?

[Anonymous Coward]

Who said anything about having a tank of liquid nitrogen laying about?

http://www.suptech.com/cryocooler.htm

Re:SQUID's next?

[DarkKnightRadick]

That's pretty hot.

Re:SQUID's next?

[oblio_one]

This is accurate, the guys above and below reporting much hotter are citing very unusual cases.

Re:SQUID's next?

[petermgreen]

You could but it would create a couple of serious complications.

1: You'd probably want the whole platter at the same temperature as a thread
2: moisture would be a menace so you would have to have the platters running in dry air. That would mean you would have to either come up with some system for drying air that went through the equalisation vent or completely seal the drive case (hard drives are not sealed because of they were air pressure changes would put huge stress on the case).

I doubt you coudl easily get the air dry enough to avoid condensation problems with a simple filter so I suspect you would either want a pressure case or a bladder.

I think you could probablly build such a drive to fit in a 5.25 inch bay and maybe a 3.5 inch bay but I very much doubt you'd get it in a laptop (which is the area where drive capacity issues bite most acutely)

Re:SQUID's next?

[DarkKnightRadick]

ah.

Re:SQUID's next?

[Anonymous Coward]

Squids are sensitive but not very directional. I don't think we'll be seeing them in drives anytime soon. Tunnel magnetoresistance is pretty sensitive in any case.

The "cosmic ray" issue is a non-issue. Errors are why drives already have extensive ECC. So long as the increase in density increases the shannon capacity it doesn't matter how high the error rate is... some amount of ECC will be optimal and will give acceptable performance.

Re:SQUID's next?

[Anonymous Coward]

Yeah that 1 or a 0 thing does seem really limited what if we did something with the strength of the magnetic field maybe with ranges then we could have more information in the signal with the minor tradeoff of increased complexity. I think I'm on to something here I will call it... analog.

Maybe other technologies as well

[mlts]

There are other technologies that I'm sure HDD makers have waiting in the wings. If areal density doesn't go up fast enough, I'm sure that HDD makers will go back to stacking platters, and we will start seeing fatter 2.5" drives. Perhaps even a return of Bigfoot drives, or double-height 2.5" drives as a new form factor. Of course, these drives will have to have some engineering done to keep performance.

I can see a full height 5.25", a monstrosity these days, but inside it would have a bunch of tiered storage with the controller doing the work and multiple caches using not just DRAM, but flash RAM, and wise positioning of data (more commonly accessed stuff closer to the spindle for example.)

This is the last resort of drive makers, but I'm sure if nothing else pans out to keep capacities growing, they will start adding platters.

Re:Maybe other technologies as well

[Grishnakh]

Maybe I'm missing something, but I don't see why they've been moving to smaller 2.5" drives anyway (except for the notebook market, where obviously small size is important), instead of sticking with larger 5.25" drives. There's a lot more area on a 5.25" platter, plus if you spin it at the same speed, the tracks at the outer edge are traveling faster and thus can be read/written faster, resulting in higher bandwidth.

The main disadvantages I can see are 1) higher materials cost, since you need a bigger chunk of aluminum, and 2) more flexing in the platters due to their larger diameter.

Re:Maybe other technologies as well

[Anonymous Coward]

Seek times?

Re:Maybe other technologies as well

[mlts]

Two disadvantages: The time it takes to move the head from the inside to the outside, and the time it takes for the drive to physically rotate the data to the head. Computers work in nanoseconds and faster. Drives have millisecond access time which is millions of times slower if one doesn't factor in caching. This is why seek algorithms and finding the best path for a drive head to pick data up from tracks is so important on a HDD.

With advances in hierarchical storage, I can see a half height 5.25" drive form factor consisting of multiple 2.5" drives and some flash memory. The HDD controller will have to be intelligent enough to figure out what stuff goes on the slow platters versus what stays on flash, and when the drive is not busy doing I/O, move things up or down the storage tiers for the most efficient way to store files. For example, WoW screenshots which are taken and forgotten end up on the outer edge of the slowest spinning platters, while a VM swap file winds up on the flash part.

Re:Maybe other technologies as well

[EvanED]

Two disadvantages: The time it takes to move the head from the inside to the outside, and the time it takes for the drive to physically rotate the data to the head.

If you take your parent's assumption that it's spinning at the same speed, then the rotational latency is independent of diameter.

But I think that's a big assumption: rotating at the same (rotational) speed means that not only do the edges wobble more, but the whole disc is under more stress. It's also generating more heat.

Re:Maybe other technologies as well

[petermgreen]

The problem is the forces involved with a spinning disc get much worse with diameter.

The moment of inertia of a cylinder (of constant thickness and density) is proportional to the fourth power of the diameter. Afaict (read: i'm trying to do calculus in my head here) the centripetal force required to hold the cylinder together is proportional to the cube of the diameter and proportional to the square of the angular velocity (spin speed).

Re:Maybe other technologies as well

[rdebath]

It is a huge assumption, back in the nineties the 3.5 inch drives won because they were big enough and had a speed that was noticeably faster than any 5.25 drive could manage.

It's the same reason that SSD drives are so popular for notebooks now. Personally, I'm also now specifying desktop machines with at least two drives (and SSD and a disk) whenever I can.

Re:Maybe other technologies as well

[Pentium100]

However, physically large drives would be great for storing information that is not accessed frequently or does not require fast access. The access times would still be much better than those of tape and their price per terabyte would be lower than that of faster HDDs and especially SSDs.

For example - I want to keep system files (and the page file) on a very fast drive, but I don't really care if my movie takes 0.1s longer to start playing. So, I could use a small, but fast drive and a slower but very high capacity drive.

Re:Maybe other technologies as well

[Joce640k]

...or just go back to 5.25" for big storage.

They'll lose the speed war to SSDs so they'll be competing on capacity. 5.25" disks will have three or four times the capacity without increasing the number of platters/heads (which is expensive).

Re:Maybe other technologies as well

[Joce640k]

The speed war is lost to SSDs. A year or two from now hard disks will be bought for capacity, not speed.

Re:Maybe other technologies as well

[Strilanc]

It's squared, not fourth power. I have no idea where you pulled that from.

Re:Maybe other technologies as well

[petermgreen]

http://hyperphysics.phy-astr.gsu.edu/hbase/icyl.html#icyl2

It depends on what values you consider constant. If you consider mass as a constant then moment of inertia is proportional to the square of the radius but if you consider density and thickness to be constants then moment of inertia is proportional to the fourth power of the radius.

IMO when talking about platters considering thickness and density as constants is more sensible than considering mass as a constant.

Re:Maybe other technologies as well

[toddestan]

The main disadvantages I can see are 1) higher materials cost, since you need a bigger chunk of aluminum, and 2) more flexing in the platters due to their larger diameter.

And 3) Having to overcome the reputation of the Quantum Bigfoot drives.

Re:Maybe other technologies as well

[Grishnakh]

And 3) Having to overcome the reputation of the Quantum Bigfoot drives.

Which was what? Sorry, I must have missed that event. The Wiki page doesn't mention any serious problems with the drives, in fact they were the first to achieve double-digit GB capacities, and their final drive, the 19GB TS, was the largest available at the time. They weren't particularly fast, though, but they were economical.

SSDs are the future

[bzzfzz]

I think a more realistic assessment is that the rate of growth in hard disk densities will decline.

We've had a recent article on the shortcomings of SSDs, but I think the maturity of hard disk technology and the minimum cost posed by the complicated mechanical design will make hard disks obsolete for most applications in a few more years. Hey, people thought 3.5" disks would be here forever, too.

Re:SSDs are the future

[Monkeedude1212]

Oh there's no doubt that they'll be obsolete eventually, the question is really just when. Will there be a big breakthrough in the next 5 years? Or will it take 20, 30, 50?

When you look at investing in things like backups - or how to keep up with competitive standards - a timeline is good to know. You might get the cheap hard drive option knowing that SSD's will have matured in the next few years - or if it's going to take a long time for them to overcome hard drives, you might invest in something you know is going to last that long.

Re:SSDs are the future

[Surt]

Let me know when you can no longer buy a 3.5" disk at fry's. Look forward to hearing from you.

Re:SSDs are the future

[Lunix Nutcase]

but I think the maturity of hard disk technology and the minimum cost posed by the complicated mechanical design will make hard disks obsolete for most applications in a few more years.

Such has been the claim for easily more than a decade and yet HDDs are still around.

Hey, people thought 3.5" disks would be here forever, too.

Since when did they leave? Floppy drives and floppy disks.

Re:SSDs are the future

[Grishnakh]

Except that no one ever uses them any more, except perhaps for weird old equipment which has built-in 3.5" drives (such as many 10-20 year-old oscilloscopes). Except for cases like that, no one uses them on PCs any more as they are completely obsoleted by 1) networks, 2) USB drives, and 3) other Flash media like xD/SD.

3.5" disks were obsolete long before people finally gave them up in favor of CD-RWs, networks and USB drives, but the industry never standardized on a replacement, and instead had a bunch of format wars with Zip disks (100 and 250MB), LS-120 disks, 2.88MB floppies, and others. Since the industry never standardized on any of these, and none became standard equipment in shipped PCs (as 1.44MB drives were), no one ever bothered buying them, and they all died out when other technologies made it feasible to not bother having removable/portable magnetic disk storage.

Re:SSDs are the future

[Jesus_666]

Not until someone starts pumping some research Dollars into that technology. Right now, SSDs are developing at a glacial pace. If they seriously want to displace HDDs they'll need four times the density, four times the longevity and one quarter the price. Or the same properties as today and one tenth the price.

SSDs are great for certain applications and less great for others. Unfortunately for the SSD makers, the main application many users have is "I need lots of cheap storage", which current SSDs are horrible for.

Re:SSDs are the future

[Anonymous Coward]

Many many people still use floppies.

You may have given them up but you are not everyone.

Re:SSDs are the future

[Pentium100]

I still use them. Floppy disks are faster than CDRW if you need to write a small amount of information (that fits on one floppy), it is also easier to make older PCs boot from them, unlike USB flash or hard disk drives. So, for example, a copy of MHDD goes to a floppy disk.

My main PC has a LS120 drive though. It can read/write regular floppies faster than the standard drive.

Oh, and all floppy disks have a hardware write protect switch, if I put a disk in a virus infected PC (to, for example, find out what hardware it has so I can find drivers for when I format the disk and reinstall Windows), I'd rather write protect it.

Re:SSDs are the future

[dylan_-]

Not until someone starts pumping some research Dollars into that technology. Right now, SSDs are developing at a glacial pace.

That's completely backwards. The drop in price per GB for SSDs has been much faster than that of HDs. In fact, if SSDs and HDs continue to drop in $/GB at the same rates they previously have, they'll be equivalent by around 2015.

Of course, that's a big "if", but it does show the progress they're currently making.

(sourced this from some graphs I found a few weeks ago. Google for "dollars per gigabyte" and I'm sure you'll find them)

Re:SSDs are the future

[Jesus_666]

Well, I'm comparing reality to the usual "SSDs are going to completely displace HDDs within the next ten years". If we want HDDs to be gone within ten years we need SSDs to be superior in most regards within the next two or three years.

Of course I'm of the opinion that HDDs are going to stick with us for a long time just like tape does - they're very good at large writes, which SSDs aren't, so people who often do large writes are going to stick with HDDs. (All of this is apart from the usual issues of storage space and longevity.)

As for price drops: I'm not sure whether SSDs can sustain this kind of growth. It's unlikely that Intel et al. are not already using fairly advanced lithography so shrinking down the flash chips isn't going to be easy. Remember that price is just one advantage of HDDs; storage space is the other. It's unlikely that we're going to see a 3 TB SSD within the next five years.

Re:SSDs are the future

[dylan_-]

As for price drops: I'm not sure whether SSDs can sustain this kind of growth. It's unlikely that Intel et al. are not already using fairly advanced lithography so shrinking down the flash chips isn't going to be easy.

I remember thinking the same thing about CPUs, RAM and HDs! I'm still seeing many stories about new techniques for SSDs being tried, so I actually think they will continue improving for a while yet.

Remember that price is just one advantage of HDDs; storage space is the other. It's unlikely that we're going to see a 3 TB SSD within the next five years.

There's already a 1 TB one available, even if it's insanely expensive, so I'll take that bet. In fact, I'll raise it: I predict there will be a 4 TB SSD for sale before the end of 2012!

Hard drive are gone, floppy style

[Anonymous Coward]

Forget hard drives, currently they are the main bottleneck in your computer, SSDs and the like are the future.

Actually we can even see now that ram is obsolete, once SSD catch up in speed (you don't even need current ram speed) why would anyone care about transfering data to ram, work on it then store it back? Just work straight on your data, gone are the days of saving, now will be the days of deleting, temporary working directory...

hard drives, they ain't part of the future me thinks.

Re:Hard drive are gone, floppy style

[Guspaz]

The fastest RAM available today operates at roughly a thousand times faster than flash (SSDs are only fast because they tend to have many channels (Intel uses 10) in order to improve performance), and RAM speeds continue to increase by moore's law. It's unlikely that flash will ever catch up, and the limitations of flash (wear) would make it completely unsuitable, even with large improvements in number of usable cycles.

What it could be useful for is as a shadow to RAM for fast hibernation support. Imagine a computer with 4GB of RAM and 4GB of flash (with a suitable degree of parallelism for speed purposes). If you do a decent job of keeping that flash relatively up to date with the contents of system RAM such that there is a relatively minor difference between system RAM and flash at any given time, hibernations could be done in under a second, and restoring from hibernation could be done at better than SSD speeds even if the computer is using a cheaper magnetic disk.

If you were smart about it, you could even resume execution almost immediately after you copied a bare minimum of data, and allow the user to interact with the system while the rest of memory is copied from flash to RAM, handling any uncopied data the user requests on the fly.

Re:Hard drive are gone, floppy style

[sexconker]

Actually we can even see now that ram is obsolete, once SSD catch up in speed (you don't even need current ram speed) why would anyone care about transfering data to ram, work on it then store it back? Just work straight on your data, gone are the days of saving, now will be the days of deleting, temporary working directory...

This is the dumbest thing I've ever read.

Re:Hard drive are gone, floppy style

[Anonymous Coward]

Between my workstation and server I have about 8TB of space in use. Now tell me exactly how I can do that with a few SSD's. By my calculations that's about 23 thousand dollars worth of SSD's (if it were even possible to put 100 of them on one computer).

They have a long ways to go my friend. Besides, my workstation with software RAID10 gets around 400MB/sec write speed and twice that on read which means it's at least as fast as one or two SSD's anyway (with the exception of seek/thrash speed obviously).

Another thing, I have tried a few SSD's myself and know many people that have tried them in development and server systems. Not a single one lasted more than a year. Extremely less reliable for development and database servers.

Re:Hard drive are gone, floppy style

[jedidiah]

If the hard disk is the main bottleneck in your desktop OS, then you need to get a new OS.

Re:Hard drive are gone, floppy style

[Anonymous Coward]

I've heard that 20 years ago when the future was non-removable optical read/write disks.

The latter are nowhere now and hardisks are still around. So are magnetic tape backups from 50+ years ago.

Re:Hard drive are gone, floppy style

[Jah-Wren Ryel]

Actually we can even see now that ram is obsolete, once SSD catch up in speed (you don't even need current ram speed) why would anyone care about transfering data to ram, work on it then store it back? Just work straight on your data, gone are the days of saving, now will be the days of deleting, temporary working directory...

This is the dumbest thing I've ever read.

You must be new to teh internets.

Re:Hard drive are gone, floppy style

[TeknoHog]

Back in the day, machines used to run straight from a ROM, with some RAM available as well. Some embedded systems still do this. In a sense we have taken a step backwards, having to copy stuff onto RAM in order to use it.

Of course, right now the only affordable choices are to use hard drives or flash with DRAM. It's not like we have to live with their limitations forever. Flash, in particular, was never really meant for frequent writing; it is a form of EEPROM. Even getting back its role as a ROM, where you could execute code directly, would be interesting.

Re:Hard drive are gone, floppy style

[Grishnakh]

The fastest RAM available today operates at roughly a thousand times faster than flash (SSDs are only fast because they tend to have many channels (Intel uses 10) in order to improve performance), and RAM speeds continue to increase by moore's law. It's unlikely that flash will ever catch up, and the limitations of flash (wear) would make it completely unsuitable, even with large improvements in number of usable cycles.

I wonder how the new PCM (Phase-Change Memory) from Numonyx will fare here.

Re:Hard drive are gone, floppy style

[Anonymous Coward]

Uh no, I don't think so, because Moores law is about the number of transistors (or other features) per dollar. And even if it was about speed that's still not true, because RAM speeds are not increasing exponentially.

Flash storage capacity on the other hand is developing faster than Moore's law at the moment. That helps to increase capacity and lower power consumption. Capacity and power consumption are extremely important factors in many applications today.

And wear is not really a big problem. To the extent that it is, it will be solved by increasing storage capacities. If you have 4 TB of flash it will wear out 1000 times slower than if you have 4 GB of flash all else being the same.

So you're not going to see a system with a 1:1 ratio of RAM and flash.

Re:Hard drive are gone, floppy style

[petermgreen]

Now tell me exactly how I can do that with a few SSD's
Well you can get 1TB SSDs, you should be able to get 8 of them in one machine relatively easilly if you buy the right case and maybe add one extra controller. The cost is indeed pretty insane though (I make it about $25K for 8x1TB drives)

Personally for desktops and servers I think a mixture is the way to go. SSD for things that get heavy random access (e.g. the OS and apps) and HDD for everything else. Even with a fairly heavy app load system drives don't generally need to be massive. Unfortunately for laptop users most laptops either only support one HDD or require you to sacrifice the optical drive to add a second HDD.

Re:Hard drive are gone, floppy style

[Nimey]

Hi, welcome to Slashdot!

Re:Hard drive are gone, floppy style

[Anonymous Coward]

If it weren't for my horse...

Re:Hard drive are gone, floppy style

[Matt Perry]

This is the dumbest thing I've ever read.

Welcome to the world of tomorrow!

Re:Hard drive are gone, floppy style

[omglolbah]

Which unfortunately would require a complete re-design of the CPU hardware ;)

Hell, your cpu doesnt even work directly on "system ram", it all goes through various buffers (L1-L3 cache etc).

For a very nice article on how the architecture works: http://lwn.net/Articles/250967/

Re:Hard drive are gone, floppy style

[Khyber]

not as fast as typical RAM but with 10^8 R/W cycle estimate (low-ball) I'd be quite happy to have one. I've bene screaming about the tech for a while on slashdot.

Re:Hard drive are gone, floppy style

[Khyber]

"So you're not going to see a system with a 1:1 ratio of RAM and flash."

Sir, I have 4GB RAM and 4GB SD Flash in my system.

What's this nonsense you speak of?

Re:Hard drive are gone, floppy style

[EvanED]

And what OS, pray tell, would you recommend?

Re:Hard drive are gone, floppy style

[IICV]

For something this retarded, you have to go further than that: I award you no points, and may God have mercy on your soul.

Re:Hard drive are gone, floppy style

[Anonymous Coward]

Back in the day, fast machines had to clone the ROM contents to RAM so the frequently accessed firmware was actually usable (int10/int13 i'm looking at you) - Bios shadow ftw!

Re:Hard drive are gone, floppy style

[Anonymous Coward]

If non-volatile memory speed ever catches up to volatile memory speed, a "working area" (i.e. what people commonly know as RAM) will no longer be necessary. This is not a dumb idea. It's a possibility.

Your post is the Unfunniest Score:5 Funny that I've ever read.

Re:Hard drive are gone, floppy style

[Arbition]

Actually we can even see now that ram is obsolete, once SSD catch up in speed (you don't even need current ram speed) why would anyone care about transfering data to ram, work on it then store it back? Just work straight on your data, gone are the days of saving, now will be the days of deleting, temporary working directory...

You must have seriously misunderstood the technology behind NAND Flash and (D)RAM. With this you may as well say that processor caches are on their way out... Which would be disastrous. Even if Flash were to reach the bandwidth and rewrite cycles of current RAM, we would still have serious problems arising from the block nature of Flash. In ram, accessing a single 32bit value is simple enough, but in Flash, you have to wade through all the surrounding data to get it (by which I mean fetch it). In fact, just to write a single 32 bit value, the block (Say, 4K, though it can be much, much more) must be fetched by the controller, the data at that memory location placed in the correct position, then the whole block written back, all while the processor waits to write a value in the slot just next to it, where it must do that cycle again. All in all, latency would make it unbearably slow, and RAM is already slow enough due to the speed of light (or electrons, you know what I mean).

Even worse is when you use temporary files. This is the single most idiotic thing I have heard. Do you even know how filesystems work? Needless to say, the filesystem overhead and the requirement for memory to access it makes it a very very poor replacement to the memory it may be intended to replace.

Forget hard drives, currently they are the main bottleneck in your computer, SSDs and the like are the future.

This is why we have caching. And RAM, And processor L1, L2 (L3) cache to make up for the slow access times. This applies to SSD's too.

hard drives, they ain't part of the future me thinks.

This is a point for debate. But there is one thing that hard drives have that SSD's do not. Density. Unlike hard drives, Flash is always connected. That is to say, that each block has a permanent circuit to access it. This takes up space on the chip, reducing the potential density. Hard drive platters are made up purely of the magnetic wells intended to store the data. This is why Seagate trialled the hybrid SSD/Magnetic hard drive (in effect adding another level of caching). One thing is certain, however. Flash/SSD is certainly preferred for mobile computing (including MP3 players) due to their lower power and physical durability.

Re:Hard drive are gone, floppy style

[Anonymous Coward]

The fastest RAM available today operates at roughly a thousand times faster than flash (SSDs are only fast because they tend to have many channels (Intel uses 10) in order to improve performance), and RAM speeds continue to increase by moore's law. It's unlikely that flash will ever catch up, and the limitations of flash (wear) would make it completely unsuitable, even with large improvements in number of usable cycles.

What it could be useful for is as a shadow to RAM for fast hibernation support. Imagine a computer with 4GB of RAM and 4GB of flash (with a suitable degree of parallelism for speed purposes). If you do a decent job of keeping that flash relatively up to date with the contents of system RAM such that there is a relatively minor difference between system RAM and flash at any given time, hibernations could be done in under a second, and restoring from hibernation could be done at better than SSD speeds even if the computer is using a cheaper magnetic disk.

If you were smart about it, you could even resume execution almost immediately after you copied a bare minimum of data, and allow the user to interact with the system while the rest of memory is copied from flash to RAM, handling any uncopied data the user requests on the fly.

You should do the math some time. For normal use an SSD will outlast a hard drive by many years and should fail in a way that allows reading of the data for the most part, unlike hard drives. They can be faster, more reliable, but they aren't yet anywhere near as cheaper per gigabyte. With that said I expect the low end to be canabilized by SSD's in the next year or two since they can be cheaper than hard drives in theory if you don't need huge size.

Re:Hard drive are gone, floppy style

[symbolset]

You've got this one nailed. The arrays of SSDs like whiptail aren't for regular servers. They're for high I/O tasks like virtual machine hosts that consolidate the workloads of dozens of servers or scores of desktops.

But when you use them for their proper purpose they can be more economical than providing physical disk for all the consolidated machines.

Buying a couple to throw at a storage bottleneck in a pinch can be handy too.

/In Seattle we have aurora now.

Re:Hard drive are gone, floppy style

[Anonymous Coward]

ROM's are slow too, that is why on PC motherboards is only a single (8 bit) Flash/Eprom which is copied to the much faster RAM at startup. For microcontrollers the slow ROM is not a problem as clock speeds are relatively slow compared to desktop CPU's

Re:Hard drive are gone, floppy style

[putaro]

Multics lives!

Re:Hard drive are gone, floppy style

[TeknoHog]

ROM's are slow too, that is why on PC motherboards is only a single (8 bit) Flash/Eprom which is copied to the much faster RAM at startup. For microcontrollers the slow ROM is not a problem as clock speeds are relatively slow compared to desktop CPU's

That was kind of my point -- at the moment, RAM is fast and everything else is slow, so we design our systems around these limitations. This, in turn, limits the range of new technologies we might want to develop.

Re:Hard drive are gone, floppy style

[Anonymous Coward]

thank you

As for sexconder:
you sir lack a lot of vision

(call it, ssd, megaholostorage or zygmatwat I couldn't care less, storage will reach very fast speed and ram won't be necessary, if you think that's dumb, technology might not be a good field for you to work in...)

Re:Hard drive are gone, floppy style

[Maarx]

I can see RAM-less computers for work / word processing / PowerPoint, but I do not forsee RAM-less computers for computationally intensive work or 3D gaming. Not within our lifetimes.

Re:Hard drive are gone, floppy style

[ImprovOmega]

I don't see it happening. At a bare minimum, cache memory will always be faster just because it's baked on to the CPU and it takes less time for the signal to travel there.

Intuition tells me that no matter how fast non-volatile memory gets, it will always be outstripped by volatile memory because you don't have to concern yourself with permanently storing it.

Re:Hard drive are gone, floppy style

[Guspaz]

You should do the math some time.

RAM is modified a lot more often than data on the hard disk, and it has to be fast too. Let's assume that, in order to match current high-end RAM speeds (GDDR5 is over 20GB/s, so I'll just use 20 for simplicity), we keep the current number of channels in an Intel SSD (10) and bump up the per-chip flash speed to reach that 20GB/s, which gives us 2GB/s (which is about a hundred times faster than what's on the market today)

To hit the 10,000 write limit of a MLC on a 2 gigabit flash/RAM chip (2gbit seems to be about the limit of current RAM chips) with theoretically perfect wear leveling would require just over 20 minutes. SLC would bump that up to 200 minutes, still not practical.

Flash for RAM isn't going to happen any time soon. The speed just isn't there, and the rewrite cycle count would probably need to be three or four orders of magnitude higher than the best we have today.

Superstition?

[PingPongBoy]

I remember when processor MHz ratings went from 566 to 600 to 633 to 667. On this disk when they achieved the 666th Gb, it wasn't good enough to report until the 667th was reached, barely squeaking over the bar.

Re:Superstition?

[Anonymous Coward]

Most 567MHz processors are labeled as 567MHz, not 566. They are all rounded properly.

Re:Superstition?

[Anonymous Coward]

I remember when processor MHz ratings went from 566 to 600 to 633 to 667. On this disk when they achieved the 666th Gb, it wasn't good enough to report until the 667th was reached, barely squeaking over the bar.

Simple rounding - 2/3 of thing gets rounded up to one whole :)

Re:Superstition?

[toddestan]

Not true. Only the 666+2/3 Mhz processor got rounded to 667Mhz. Every other model truncated the 2/3 and went with 266, 366Mhz, etc. For example, look at the following list. (the Celerons had the most 66/67 processors due to Intel sticking with the 66Mhz bus speed for so long, but it holds true for the Pentium II/Pentium III also)
http://en.wikipedia.org/wiki/List_of_Intel_Celeron_microprocessors#.22Covington.22_.28250_nm.29

Just use DNA

[WillAffleckUW]

The limits on recording may not approach the Dirac effect, but there is a heck of a lot of information, especially with siRNA, miRNA, mRNA, and other modifiers that allow multiple usage of DNA strands to adapt and record.

Music-playing microbes could store more data than there are songs in the world from the beginning of time until now. In each microbe.

Re:Just use DNA

[faragon]

Ok, and what about random access time?

Re:Just use DNA

[WillAffleckUW]

RNA splicers can run on multiple copies of DNA segments at the same time, churning out segments and proteins in response to the events.

You do realize your OLED TV is a biological device, don't you?

Re:Just use DNA

[faragon]

I don't doubt about organic storage capabilities, I was asking about random access time, capisci?

Re:Just use DNA

[WillAffleckUW]

I just told you that multiple readers can work on multiple strands - how do you think you "think", by the way ... it's a biochemical process and your nerve cells are what let you "hear" and "see".

Buffering is not a big deal with access. You are already limited by your perception limits.

Re:Just use DNA

[maxume]

Are you serious? OLEDS are about as biological as a chunk of graphite.

Re:Just use DNA

[WillAffleckUW]

graphite can be produced by biochemical means too.

You have no idea how advanced things are ... here at the UW we hold patents on biofilms for solar cells, adaptive taggants that target cancers and glow in the dark to aid in surgical removal of cancerous tissues, and we even grow living tissue.

All DNA is is just a storage system for information that is read by organic mechanisms. It even has self-repair mechanisms and adaptive repair and most biochemical pahtways have 2-3 backups in case stuff happens. A lot of what people used to think was 95 percent noise in DNA is actually overlapping instruction sets activated by silencing and other RNA sequence readers and activators.

You're thinking old tech. Wetware is where it's at.

Re:Just use DNA

[maxume]

So you are saying that OLEDs illuminate as a result of some biological process, that it isn't just some electrons activating some crystals?

And I'm not sure why growing living tissue is a big deal, I've been doing it all my life (of course, you probably mean growing differentiated tissues in a dish, but given how willing you are to blur lines, why not join in on the fun?).

Re:Just use DNA

[WillAffleckUW]

actually, as you know, the UW is one of the primary research universities, especially in the fields of genetics, biochemistry, microbiology, nanotechnology and many other fields. I'm referring to growing actual surgical replacement tissues, to modifying actual genetics - cutting edge techniques - pick up any issue of Science or Nature or check PloS.

My point is that you're suffering from the same lack of foresight that we had back in the 80s about magnetic memory and disk storage. What we think of as "storage" back then and what we think of as "storage" now continue to shift their frames of reference. Most data and information storage techniques continue to evolve. We know that data storage and information storage using DNA and other biochemical processes is not just possible, but provable and can be implemented. That we artificially restrict our concepts of "how" we store and access data, and protect it from degradation/error, and encyrpt it, only shows our own limitations, not the actual limitations of science.

All your data is belong to us.

Re:Just use DNA

[maxume]

Actually, I don't really know anything about UW. Sorry.

And I don't know why you are going back to storage over and over, I'm tiresome-internet-guy picking apart your silly statement about the organic molecules used in OLEDs making them biological, not someone objecting to the idea that it may be possible to create practical systems for storing information in DNA (or some similar thing).

(And I get that DNA or RNA is used by every known living organism to store information, I mean stuff like movies)

Re:Just use DNA

[Rakishi]

And nerve cells don't use DNA for storing information, so your point is what again?

Re:Just use DNA

[Anonymous Coward]

It was random, wasn't it?

Re:Just use DNA

[Anonymous Coward]

Music-playing microbes could store more data than there are songs in the world from the beginning of time until now. In each microbe.

Ooh, nice, self reproducing piracy! Take a common bacterium, alter it in some way that will be easy to see in a microscope, and then fill it with music DNA.

Or the Murphy version

[Anonymous Coward]

When the only tool you have is a hammer every problem requires a screwdriver.

Re:Or the Murphy version

[Conditioner]

when the only tool is to get hammered. then everything you see gets nailed

waste.. won't ...fit

[h00manist]

Drive full. It's packed in here dude. Please release some space. Like those potential-evidence 121,000 scanned federal court transcripts from 1972 you found in grandpas basement, but havent read yet to decide if it belongs on some icelandic website... Nobody's gonna read it all.

Re:waste.. won't ...fit

[jedidiah]

Grandmas vacation video in AVCHD will quickly dwarf all of that.

Re:TEA is unreadable.

[h00manist]

This is the internet. Facts, spelling, and concepts are all optional.

Nah. TEA is unreadable. Especially the leaves.

Oh noes!!

[countertrolling]

All that effort spent trying to keep my drives cool gone to waste... Can't we move beyond magnetic.. into subspace or positronic or something?

I thought SSDs were at their limit...

[Anonymous Coward]

Funny, a week ago there was an article on /. which said SSDs won't replace hard drives in the near future because of their limited storage density...

That's about 739Gb/sq more than you need

[Anonymous Coward]

All that computing power, all those Gigabytes, the big wide screen, and what do you do with it - write a classic novel; edit your latest film; analyse that huge data set you accumulated through your amateur observatory? Nope. You read your email, watch films and play games. Sure you do a bit of techie stuff but that is for your boss.

Never has so much technology been put to so little use as it has during the last fifty years.

The complete works of Shakespeare could fit into 20Mb (or less). If you think you need more than that, you are over optimistic about your likely productivity.

More than feasible

[davmoo]

From the summary:

It's great that we can now store 2 TB on one hard drive and that 3-TB hard drives are already feasible.

3TB drives are already well past "feasible". Seagate has one for sale in the form of the STAC3000100 FreeAgent GoFlex Desk. Its an enclosure with a single SATA 3TB hard drive. The reason its currently only available as an external drive is because most motherboards will not support a boot drive that large, hence not a lot of reason to offer it as an internal yet.

Re:More than feasible

[Fumus]

most motherboards will not support a boot drive that large

Do you mean it won't boot if the main partition is 3TB or if the whole drive is 3TB? Because if it's the first, then no sane person has just one partition on a hard drive that big so it's a non-problem.

Re:More than feasible

[rockNme2349]

So you're saying 2 TB aught to be enough for anybody?

Re:More than feasible

[davmoo]

Its my understanding, and someone correct me if I'm wrong, the whole drive will not work with most current hardware. It has to do with LBA, 64-bit addressing, yadda yadda yadda.

Re:More than feasible

[guruevi]

It's not that motherboards won't support it, it's that Windows (even 7) won't support it. You CANNOT boot Windows from a disk with GPT. You also CANNOT boot Windows (7) on most EFI systems.

Re:More than feasible

[rdebath]

Ooooh. Can of worms time.

The problem is that a drive of two real terabytes (not marketing terabytes) is one 512byte sector too large for a normal MBR partition table. So you have to switch over to something else; Windows uses a GPT style partition table.

Unfortunately the committee who designed GPT were dumb (Some of the members must have been smart though, you can easily work around every dumb choice I've seen). A current BIOS doesn't know anything about partition tables and it has no problem with drives up to (IIRC) 64 petabytes. The committee decided that the BIOS should contain a boot manager that understands GPT structures. The BIOS makers don't want this problem; their stuff works fine, some people are already using RAID devices with far in access of 3TB and every OS that can handle a 3TB drive can boot off it ... except Windows.

So it's SNAFU by Microsoft. The weird thing is this time they did it by following the written standard and ignoring the defacto standard ...

Re:More than feasible

[jonbryce]

It is not available as an external drive because it is a 5 platter drive which is too fat to fit inside standard sized drive bays.

Re:More than feasible

[Anonymous Coward]

I was under the impression that most EFI systems were Macs. In which case: http://support.apple.com/kb/HT3986 You can boot windows 7. Thanks for playing.

Re:More than feasible

[Anonymous Coward]

Plus, 3TB? Come fucking on. 1TB 2.5" drives have been around for ages and you can fit four of those in the space that 3.5" takes.

739 Gb/sq.in.

[Mojo66]

739 Gb/sq.in. equals

106416 Gb/sq.ft.
957744 GB/sq.yd
2966707814400 Gb/sq.mile

It also equals
1.145452290904 GB/sq.mm
114.5452290904 GB/sq.cm
1145452.290904 GB/sq.m
1145452290904 GB/sq.km

Re:739 Gb/sq.in.

[finity]

How many libraries of congress/libraries of congress is it?

Re:739 Gb/sq.in.

[Gunstick]

I don't know if "library of congress" is a surface measure. I only know football fields. But maybe someone knows the size of the Library of Congress in terms of football fields so we could make the conversion.

Where do you back it up?

[Alcoholist]

"Where do you back it up?" is what I ask my customers. If you buy a terabyte sized hard drive, what's your solution if it fails? Presumably you bought it so you could store zillions of pictures, MP3s and movies on the thing... how badly will your day be ruined if it fails?

Drives that big, you buy them in pairs, one mirrored to the other.

Re:Where do you back it up?

[EvanED]

Drives that big, you buy them in pairs, one mirrored to the other.

If my "mirrored" you mean "RAID 1", I would say that barely counts as a backup. There are essentially three or four substantial threats for why you need a backup, and RAID 1 protects you against just one of them. (If you're counting, the four threats are (1) drive failure, (2) your power supply committing murder-suicide and taking out your drives, (3) your house burning down or computer being destroyed (you can combine 2&3 if you want), (4) software corruption.

The first rule of backups is "make backups". The second rule of backups is "make backups". But the third rule of backups, if you ask me, is "RAID isn't backups". (The fourth rule is perhaps "check your backups to make sure you can restore from them.")

(What you should do is buy in pairs, put one in an external enclosure, and periodically sync them. Keep the second drive unplugged from everything when it's not actually being synced.)

Re:Where do you back it up?

[spitzak]

I don't think the GP meant RAID when saying "mirroring", he actually meant "make 2 copies of your data, one on each disk".

I agree this does not protect against a disaster that affects both drives because they are in the same box.

Re:Where do you back it up?

[White Flame]

1) Do you need to back up absolutely everything? Are many of your giant files just ripped from your DVD collection, or re-downloadable? I've got a 1TB data drive and use my "old" 500GB for backup, and it's got plenty of room for the portion that actually needs backing up.

2) They're cheap. A 1TB drive is like $60 nowadays. Getting a drive smaller than that will probably be higher bytes-per-dollar, so might as well get two of the same if you don't have an older one handy. And like the others have said, mirroring is NOT a backup solution. Many classes of problems and user errors will simply reflect to both drives and leave you screwed.

Re:Where do you back it up?

[Pentium100]

LTO4 can store 800GB on a single tape and LTO5 can store 1.4TB, so you only would need a few tapes to back everything up.

Or you could buy another hard drive, place it in an external enclosure and back up to that.

Re:Where do you back it up?

[rdebath]

Except, it can take quite a while to sync two drives of this size. So you will probably find that the second drive spends at least half it's time sitting next to the primary drive.

You actually need THREE drives so one of they is always a safe backup.

The actual rules:

1. Make a Backup.
2. Make it safe.
   eg: offsite
3. Keep it safe.
   eg: THE backup must stay offsite, it only comes back when it's not THE backup.

Re:Where do you back it up?

[arth1]

There are essentially three or four substantial threats for why you need a backup, and RAID 1 protects you against just one of them. (If you're counting, the four threats are (1) drive failure, (2) your power supply committing murder-suicide and taking out your drives, (3) your house burning down or computer being destroyed (you can combine 2&3 if you want), (4) software corruption.

You forget the perhaps most common one: "Ooops, where did this folder/file go?"

The first rule of backups is "make backups". The second rule of backups is "make backups". But the third rule of backups, if you ask me, is "RAID isn't backups". (The fourth rule is perhaps "check your backups to make sure you can restore from them.")

An untested backup is fairly similar to no backup at all.
I can't even count how many times I've heard people who needs help because they can't restore their backup.
Usually this is because the backed up device needs proprietary drivers, which the standard restore media lacks. Sometimes that's easy to add, sometimes not. But when you need something restored isn't the time to do that. If you haven't done test restores, you can't count on your backup.

The second largest problem with untested restores is probably limitations with the backup software that affects the operation of a restored system. It can be that the backup software doesn't preserve advanced file attributes or permissions. Or, for NTFS, it only restores the "long" file name, while letting the OS create a new short DOS 8+3 file name, in effect rendering software using the short paths unusable after restore. (Like Microsoft Office, which will cease to work if "Microsoft Office" enumerated to MICROS~1 at install time, but became MICROS~3 during restore.)
Or that it can't handle sparse files, so your database that takes up 500 GB before restore would require a 5 TB drive to restore to.

The third big problem is that the backup doesn't include what the user wants restored. This is often due to "smart" backup software, which might back up "My Documents", but not, say, "C:\Legal\". Or it might back up just files with certain known extensions. Or it might exclude files above a certain size (yes, there is "backup" software out there that does this by default).

Then there's the big problem with out-of-date backups. If the user has to manually change CD-RWs, backups will be taken once, and two years down the road when the user needs a restore, the files are too old. Or the backup process starts failing after a patch Tuesday, but no alerts go out, and the user thinks everything is hunky-dory. Until he deletes a file by accident and needs it restored.

1: Back up.
2: RAID is not backup.
3: Until you've verified restore, you have no restorable backup.
4: Until you've verified alerts, you have no current backup.
5: "Smart" backup programs aren't.
6: Any backup requiring manual interaction will be out of date when you need it.

Re:Where do you back it up?

[dylan_-]

If you buy a terabyte sized hard drive, what's your solution if it fails? Presumably you bought it so you could store zillions of pictures, MP3s and movies on the thing... how badly will your day be ruined if it fails?

Some things stay the same! From Slashdot, 1998:

IBM announces a 25 gigger

Hardware Posted by Hemos on Wednesday November 11, @10:11AM
from the why-i-could-put-3/4-my-cd-collection dept.
Booker writes "So IBM announces a 25 gig hard drive... does the world need this yet? Unless this is in a RAID, would you really want to trust 25 gigs on a single drive? What would you use this for? 400+ hours of MP3s comes to mind... "

Re:Where do you back it up?

[Pinback]

The tape makers are working on adopting perpendicular reporting as well. Although with GMR type 2 heads.

Native capacities of 35TB per tape are expected. But the required tape drives will not sell for 149$ on Newegg.

Re:Where do you back it up?

[steveg]

No sweat. An LTO4 tape is fairly inexpensive, on the order of $50 each. A bargain, considering the capacity and ruggedness.

Only...

A drive will set you back somewhere between $1500 and $3000. Is it worth it? Well, yeah, maybe. Definitely, for a business. Not practical for the average home though, and that's where most of these TB+ drives are going. And most small businesses won't spend that kind of money, even if they should. *I* probably wouldn't, if it were my money.

LTO tape is a far safer solution than an external hard drive, but it's not more *practical* for most people.

Re:Where do you back it up?

[Pentium100]

I bought a used LTO2 tape drive for ~$190. While the tapes are more expensive per GB and only 200GB, the drive was affordable and since I mostly use tapes for archiving and record whole tape at once, 200GB is enough for me for now. When 200GB starts being too small, I'll probably be able to get a LTO4 tape drive for $200...

HP diagnostic software says that my drive still has ~99% of head life left.

Re:Where do you back it up?

[BlackHorse]

Mod points I lack, but you deserve.

Seagate is already selling a 3TB drive

[lazn]

Seagate is already selling a 3TB drive.

http://www.pcworld.com/article/200031/big_seagate_3tb_drive_ups_storage_ante.html

bleh

[buddyglass]

For a good while now the size of drives has been mostly meaningless to me. I don't store any movies or music. My current XP installation, with MS Office and Eclipse, takes up about 10 GB. I'm much more interested in "fast" than I am "big".

Re:bleh

[Joce640k]

The Intel X25-V is just what you need...

Re:bleh

[buddyglass]

If I were to buy/build a new system today I'd want an OCZ Vertex 2. Thing is, I'm fairly paranoid about quality issues when it comes to SSD manufacturers not named Intel. And Intel's best offering just doesn't measure up w.r.t writes. Since I'm "sort of" in the market right now, I may wait until this Christmas and pick up Intel's successor to the G2. Which, considering the year-old G2 is still very competitive (at least with respect to reads), should totally blow away everyone else's offerings.

down to eg -100C ?

[Anonymous Coward]

down to eg -100C or whatever is low enough to result in superconductivity. .. a quick goog gives 139 K or -134.15 degrees Celsius

I wonder how long a 3TB drive takes to defrag...?

[Assmasher]

(Presuming a large number of files of course...) Hehe.

Long term storage?

[Anonymous Coward]

And which one should I buy for long(er) term storage and/or archival purposes?

I'd think that there would be a market for drives that have larger magnetic domains. I.e. less storage per area on the platters. Think 5.25" drives with 500 GB capacity or somesuch. I'd buy those, if they where made for long term storage.

Disks aren't beaten yet

[valen]

  What about Shingle recording ? There are prototype drives already built that can near 100TB on a single device. It has limitations, of course, like 2GB+ block sizes, but it'll get interesting. Give it 3-5 years (as usual, for mythical tech that might ship someday).