SSD Annual Failure Rates Around 1.5%, HDDs About 5%

Lucas123 writes "On the news that Linus Torvalds's SSD went belly up while he was coding the 3.12 kernel, Computerworld took a closer look at SSDs and their failure rates. While Torvalds didn't specify the SSD manufacturer in his blog, he did write in a 2008 blog that he'd purchased an 80GB Intel SSD — likely the X25, which has become something of an industry standard for SSD reliability. While they may have no mechanical parts, making them preferable for mobile use, there are many factors that go into an SSD being reliable. For example, a NAND die, the SSD controller, capacitors, or other passive components can — and do — slowly wear out or fail entirely. As an investigation into SSD reliability performed by Tom's Hardware noted: 'We know that SSDs still fail.... All it takes is 10 minutes of flipping through customer reviews on Newegg's listings.' Yet, according to IHS, client SSD annual failure rates under warranty tend to be around 1.5%, while HDDs are near 5%. So SSDs not only outperform, but on average outlast spinning disks."

Comments

News at 11

[Anonymous Coward]

An intel apologist supporting intel? Shocking!

Poor statistics

[Anonymous Coward]

"client SSD annual failure rates under warranty tend to be around 1.5%, while HDDs are near 5%"
So they are less likely to fail early in their life.

NOT:
"So an SSDs not only outperforms, but on average outlast spinning disk."

This is completely unsubstantiated by the evidence provided.

Re:Poor statistics

[MickLinux]

That said, my memory was that some reported on Âlashdot that you can force the failure of an SSD by powering it down in themiddle of a write, then powring it up, causing it to go into chkdsk, and finally powering it down in the middle of chkdsk. Which is not too unlikely an occurrance. If you wanted to decrease the user failure rate, you might hook it upyo a supercapacitor.

Re:Poor statistics

[Anonymous Coward]

Preferably one with spillchick.

Re:Poor statistics

[MikeBabcock]

One of the few benefits of a spinning platter is that they can briefly generate their own juice when the power goes out.

Re:Poor statistics

[Anonymous Coward]

There's another factor they don't take into consideration - when the drive fails, in which condition will it be afterwards. I had multiple HDDs fail on me in my life, and the most common effect was inability to read a bunch of sectors. It damaged the file system and several files, but in most cases I could still mount it read-only and recover most of the stuff from it unscathed. Just a few weeks ago I had an SSD failure (OCZ Vertex3). I was working and the drive just suddenly died. Without a warning, and of course without any screeching noises. I noticed because a couple of applications crashed and could not be restarted afterwards. While the drive was still mounted, I saw that about half of directories on it became inaccessible, or disappeared. After shutting down and attempting to mount this drive elsewhere for rescue, I realized that the FS was damaged beyond any recognition - half of the sectors were unreadable, I could not recover a single piece of data from it. Yes, I had backups but as usual not necessarily the freshest, and I had to reinstall everything. At least gave the opportunity to switch to a fresh 64-bit install after I've been using my old install continuously since 2004.

Re:Poor statistics

[BancBoy]

One of the few benefits of a spinning platter is that they can briefly generate their own juice when the power goes out.

As many of us do, when the power goes out...

Re:Poor statistics

[Anonymous Coward]

Don't SSD's usually have much shorter warranty periods than HDDs? For example, if the SSD warranty is 1 year, whereas a HDD warranty is for 10yrs, and one takes into account the fact that the probability of failure increases exponentially over time, then this would mean the chance of an SSD dying within the warranty period of the HDD are exponentially higher than 5%...

Re:Poor statistics

[smash]

Not only that, if you add enough SSDs to store the same amount of bytes as a 4TB hard drive you either need to RAID them or you will have cumulative failure rate far higher than the HD failure rate, which you could mirror (or hell, RAID6 it) for way less money.

Re:Poor statistics

[ArcadeNut]

No.

I just picked up two SSD Drives that have 5 year warranties. I also picked up two Segate HDD's that only have 2 year warranties.

Most HDD's are 1 - 3 years. I have several Segate drives that are 5 year also.

Re:Poor statistics

[hairyfeet]

I would just add the whole thing ignores that big old rotting elephant in the room which is HDDs? I have found that in damned near every case, not all but most, will give you PLENTY of warning before it goes completely tits up whereas the SSD? One day its working and the next....nothing. No warning, no noise, no indication at all that there was a problem just...poof, buh bye data. This is also ignoring the fact that if the circuit board fails in a HDD you can swap one out for the same model and get it back in most cases, at least long enough to get the data, the SSD? Hope you are good with a soldering iron and a chip reader and I have heard even then its unlikely.

I may be just a little country shop guy but when my gamer customers have all experienced multiple failures when it comes to SSDs, and these guys don't go cheap, sorry but ATM I still don't trust it. I tell folks if they want an SSD don't have anything on it they would feel bad if they lost, now does that mean there aren't still uses for SSDs? Of course not, for one thing if you have a laptop where most if not all of your data is in the cloud? Knock yourself out, just make a weekly disk image so you can re-image when it goes tits up and you are golden. I also have several customers that have bought either hybrid drives or that Sandisk caching drive for Win 7 and in both of those cases they have seen pretty big speed boosts while not having to worry because if it dies all you do is go back to HDD speeds as it is just a cache.

Oh and one final thing....its gonna get worse. its common knowledge that with each shrink the number of writes goes down and the number of failures go up and with all of the major chip companies seeming to only care about how many bits they can stuff per nano-meter? The failure rate WILL get worse, you can count on it. Its too bad that SLC is so insanely high as those seem to have lower failure rates than MLC but as long as all the companies care about is getting that GB number up at all costs its really not gonna be getting better, its gonna be getting worse.

Ironic that they talk about how supposedly high HDD failure rates are when I cleaned out a how drawer of them before moving into the new place, we are talking drives going back to Quantum Fireballs in the 200Mb size, yes Mb not Gb, and they all fired up. granted some of them were noisy as hell but I could still get files off of them while not a single one of my gamer customers have their first SSD, they are all dead. yes i know its an anecdote but I'm not the only one that has seen this, coding horror calls SSDs the hot crazy scale as you trade red hot performance for crazy failure rates. Call me old fashioned but I think I'll just pick upa caching SSD and keep the 5Tb in spinning rust, thanks ever so Intel.

Re:Poor statistics

[Sir_Sri]

The way I configure SSD's is as a OS/boot drive, and then I write all user data off to a RAID with traditional HDD's.

The simplest way is a SSD for windows/linux and then put your user directories on a RAID1 of 1TB drives, and then backups from there.

Re:Poor statistics

[hedwards]

The other problem is the write cycle limit on the SSDs, that may or may not be an issue depending upon how you use your computer. But, for those of us that regularly recompile the OS and kernel, an SSD isn't going to stand up to that for very long.

Oh, and SSDs are fast, but they're still tiny. I had to replace a HDD recently and I was able to get a 1TB disk for $75 including shipping, and it's a fairly fast disk as well. I could never have afforded that much capacity with SSDs.

Re:Poor statistics

[Zemran]

I would not "force" a failure on an expensive piece of kit and the only people I know that do are researchers. You would need to be running an early file system to do that.

Re:Poor statistics

[girlintraining]

I have found that in damned near every case, not all but most, will give you PLENTY of warning before it goes completely tits up whereas the SSD?

Yeah, sure, okay. If you're sitting next to your computer, then yeah, maybe you notice. How about the hundreds of millions of drives that are sitting in a rack somewhere, and will only see a human being twice: Once when it gets installed in the rack, and then only when it stops working for whatever reason and a tech is sent out to replace it.

The "it made a funny noise first" line item is a joke either way. This is like saying "Well, I prefer diesel engines because they make more noise when they die." Hookay. Yeah.

I may be just a little country shop guy but when my gamer customers have all experienced multiple failures when it comes to SSDs, and these guys don't go cheap, sorry but ATM I still don't trust it.

I may just be a Ferrari repair shop owner, but when my car owners have all experienced multiple failures when it comes to ceramic brakes and high end engine components, and these guys don't go cheap, sorry but ATM I still don't trust it.

Now do you see how utterly ridiculous that sounds? High performance almost always means less robust. That graphics card you just plunked over $200 on? It's operating temperature is so high from the current being pumped through it that it's literally cooking itself at the molecular level from the moment you plug it in -- it's called electromigration, and in three to five depending on how often you use it, it's going to shit itself. But that's okay... because in two years, you'll be spending even more on a new one.

Ironic that they talk about how supposedly high HDD failure rates are when I cleaned out a how drawer of them before moving into the new place, we are talking drives going back to Quantum Fireballs in the 200Mb size, yes Mb not Gb, and they all fired up. granted some of them were noisy as hell but I could still get files off of them while not a single one of my gamer customers have their first SSD, they are all dead.

Yeah, and? How many gamers are still using their 200Mb Quantum Fireballs in an actual computer? I know it's a common geek past time to see what kind of antiquidated hardware you can pull out with your friends... that old parallel port Zip drive, or floppies the size of your head... and yeah, it's fun to talk about to show you had IT chops before the person you're talking to was even a glint in daddy's eye... but that's the only value they have.

Nobody's coming up to me and asking for an AT command initialization string for their modem -- AT&F&C1&D2S95=55 in case you were wondering -- because it's not a technology very many use anymore. Yeah, I can dig out an old 2400 baud modem and get it working... but that doesn't mean 2400 baud modems are superior to cable modems that "have a higher failure rate".. and so, you know... I don't know if I trust such 'new' technology.

Now, get off my lawn.

Re:Poor statistics

[icebike]

You might not do that, but your failing power supply might do it for you.

Re:Poor statistics

[icebike]

Yeah, sure, okay. If you're sitting next to your computer, then yeah, maybe you notice. How about the hundreds of millions of drives that are sitting in a rack somewhere, and will only see a human being twice: Once when it gets installed in the rack, and then only when it stops working for whatever reason and a tech is sent out to replace it.

Hmm, my drives send me emails when they start having problems. (And having gotten one of these emails a few years after setting up the drive initially, I was shocked to find it the email arrived in plenty of time. I pleasantly surprised to find the drive and all data still intact, and had time to swap a replacement into the raid).

Why don't you find out how this is handled by people who actually have hundreds of drives to deal with.
If you let them fail before servicing them you are doing it wrong.

Look into: man 8 smartd

Re:Poor statistics

[nedlohs]

Because Linus, who apparently uses SSDs, would never regularly compile a kernel or anything like that.

Re:Poor statistics

[Mashiki]

Really? Odd that I can buy SSD's in a 1.5-3TB flavor these days, they're expensive as all hell, but I can buy them. They come in PCI-e and SATA flavors. And really at that point, you're running with a mirror or shadow backup, or something anyway. Besides, if you're using a single drive like that, you're at a single point of failure at both the consumer level and at the enterprise level. But let's be honest, you can't beat good backup practices into anyone. As much as you try, and all that.

Re:Poor statistics

[Zemran]

A power failure is very different to an SSD failure. Any drive can normally cope with a power failure. You have to create the situation whereby such an event will cause damage to the actual drive.

Re:Poor statistics

[Miamicanes]

The warranty period for a SSD is almost meaningless, because most SSD failures involve instant unrecoverable data loss by a drive that technically isn't "broken" in the legally-binding warranty sense, since you can "fix" it by erasing everything and reloading the virtual gun for another round of Russian Roulette.

Re:Poor statistics

[smash]

As I said, for far less money you can get redundant storage in the multi terabyte range. My point is this: if i need to store x amount of data in a reliable way, the money i save vs. SSD on buying a spinning disk means I could buy several of them to mitigate the increased failure rate vs. SSD and far exceed the reliability of an SSD for the same money.

Re:Poor statistics

[countach74]

And his SSD failed. What is your point, exactly? :) Seriously though, from the reading I've done on it, it takes a *ton* of read/writes to blow through a decent SSD's limits.

Re:Poor statistics

[zAPPzAPP]

This sounds like a very specific problem with a certain firmware to me.
It's not an inherent problem with the SSD technology.

Re:Poor statistics

[Anonymous Coward]

Some data corruption on power failure is pretty common on cheaper SSDs that don't have a supercapacitor (or battery) for power failure protection. In general, you can't "kill" the SSD that way -- the worst failures I've seen with that locked the drive read-only, and left all but a couple of MB readable... the physical hardware was fully recoverable once I backed up the data, via a SATA secure erase. That happened a couple of times on older Crucial m4 drives; I think it's fixed to be less likely in newer firmware.

Re:Poor statistics

[Anonymous Coward]

I agree with you regarding poor statistics.

"Yet, according to IHS, client SSD annual failure rates under warranty tend to be around 1.5%, while HDDs are near 5%. So an SSDs not only outperforms, but on average outlast spinning disk."

I believe that its also normal that HDD have a higher rate of failure since there are more poeple with hdds than ssds. The more you have, the higher the % will be.

Re:Poor statistics

[mister_handy]

See also this paper/presentation: https://www.usenix.org/system/files/conference/fast13/fast13-final80.pdf https://www.usenix.org/conference/fast13/understanding-robustness-ssds-under-power-fault

Re:Poor statistics

[thegarbz]

No they don't. Standard warranties are 5years on SSDs. Some of the really cheap ones like Samsung's bottom of the line drive has 3 years.

No HDD manufacturer offers 10 year warranties on consumer harddisks that I am aware of. In fact it was only last year that Seagate and WD announced they were dropping the warranty period from 5 down to 3 years.

Re:Poor statistics

[istartedi]

And while scanning the SMART data is a nice start... you aren't going to get an e-mail when a branch office's first floor is under five feet of water

Ummm... in that case I think you'd get a phone call from a human.

Re:Poor statistics

[Anonymous Coward]

I tell folks if they want an SSD don't have anything on it they would feel bad if they lost,

If you have data you would feel bad if you lost it, there's only one correct thing to do:

Make backups.

Re:Poor statistics

[moronoxyd]

I believe that its also normal that HDD have a higher rate of failure since there are more poeple with hdds than ssds. The more you have, the higher the % will be.

So... if I compare the mortality rate of a group of 1000 people with of a group of 1000000 people the mortality rate of the bigger group is bigger?
Or the ratio men vs. women?

Failure rates are independent of the number of units in the group, as long as the sample is big enough. That's why we use % instead of actual numbers.

Re:Poor statistics

[ssam]

Was SMART showing anything before the failure?

Re:Poor statistics

[TheRaven64]

I've had three hard disks fail, none of them was in a state where any data was readable. My university computer lab at the time had about half of the drives fail within 18 months of purchase, all with the same failure mode. So I guess you were luckier. I've not had any SSDs fail yet, but I only have one and it's only about 2 years old, so there's still a lot of time...

Re:Poor statistics

[jaseuk]

Yes, but. Bad firmware that can be fixed by flashing wouldn't generate a warranty return. Yet the customer has still lost all their data.

We have a lot (around 20 out of 100 drives in just over a year) of SSD firmware issues (resulting in lost data), but only one for destruction (we don't return drives).

That said, I have 288 drives in a traditional array (between 2->5 years per shelf), we replace under warranty 3-4 a year. This is still a very small sample.

Jason

Re:Poor statistics

[TheRaven64]

Hmm, my drives send me emails when they start having problems. (And having gotten one of these emails a few years after setting up the drive initially, I was shocked to find it the email arrived in plenty of time. I pleasantly surprised to find the drive and all data still intact, and had time to swap a replacement into the raid).

There was a paper from Google about 18 months ago that showed that, while SMART errors do usually mean the disk will fail, lack of SMART errors does not mean that the disk is not about to fail. In a large number of cases, the drives failed with no warning.

The noise is also something of a red herring. I have a laptop whose drive started to make slightly worrying noises about 2 years ago, but which hasn't died (yet). I've also had several hard disks die with no warning at all. Sometimes, the failure correlates with hearing an odd noise first, but often it doesn't.

Re:Poor statistics

[Anonymous Coward]

Try to hook up one of those old drives and use it. They are probably from the era before all HDDs had wear leveling. Place a swap partition on it or so, it will wear out in a week.

Re:Poor statistics

[PRMan]

I just had a WD 3TB Green HDD fail instantly yesterday in the middle of a compile. Apparently the motors fail on these and there's little you can do to get the data off it (I lost a day's worth of data and it would take longer than that to recover it). So they don't always warn you and the failure can be instant and unrecoverable.

I have had SSDs fail when it's too hot, but continue to work afterward (data is gone, but drive still works). That could be a big problem in gamer rigs.

Re:Poor statistics

[aliquis]

Also there are multiple memory techniques used in the SSDs so reliability will likely differ a huge bit.

Also a 256 GB SSD would cost me more than a 3 TB HDD. Get 3 TB of storage space and you need 12 of those, costing what? 16 times more?
Guess it's not fair to ask what will break down first, one of 12 SSDs or a HDD because it's also 1/12 as much which break down.. But I thought I was into something there =P, get 16 HDDs and RAID them all and they will likely offer more reliable data than the 12 SSDs =P

Re:Poor statistics

[Anonymous Coward]

The big old rotting elephant carcass in your argument is, as usual, that you haven't thought things through. However, unless you disingenuously want to compare apples to motorcycles, the relevant issue isn't that HDDs usually give you some warning, but rather that sometimes it does not, or that the warning signals might for whatever reason not reach you. The question is; Are these cases more frequent than 1.5%?

YMMV, but I'd answer that with yes.

Re:Poor statistics

[nkuehn]

Absolutely true- We've had a few SSDs fail in my company (much higher rate than HDDs, but that's a different issue) and not a single of them failed gracefully. Due to internal reorganization algorithms of the SSD the filesystem's behavior to "mark bad blocks" ran into nonsense. Effect: not even a bit recoverable. New policy for every machine that has the space for an additional HDD ist to either use the SSD as a Caching layer only or to do a full image backup of the SSD every night (SSD->HDD local on the machine, ideal for laptops etc).

Re:Poor statistics

[Anonymous Coward]

You're going around in circles. This particular thread happens to be about the situations where such an event will cause damage to the drive. In particular, a power failure during chkdisk apparently does the trick.

Re:Poor statistics

[Anonymous Coward]

In fact it was only last year that Seagate and WD announced they were dropping the warranty period from 5 down to 3 years.

That seems reasonable. If a drive lasts 3 years then it probably isn't defective, right?

Re:Poor statistics

[Anonymous Coward]

It's more that if you have 2TB on a drive and 500GB on another, you're going to get more opportunity for failure when you're using the full space for the larger drive.

There were theories that soon RAID would be pointless because the size of a disk was going up vastly quicker than the uncorrectable error rate was going down, hence there would be a case where the rebuilding of the RAID of a single failed drive would be expected to get an uncorrectable bit error before finishing, requiring a new HDD and a rebuild, which would be expected to fail similarly.

And as another poster pointed out, this was failures under warranty. Since there are no moving parts in an SSD, mechanical failures are not possible and any electrical problems are found and cleared at the factory for both drive types in most cases, you would expect no less than warranty failures on HDD would be quite a bit higher than SSD.

But a HDD that lasts 2 years longer than warrantly is likely adequately built with no manufacturing defects and will therefore work happily for decades. An SSD will have the same 1.5% chance each warranty period (with an increase based on heavy read-write cycles over time, this COULD be assumed to be congruent with mechanical wear on a HDD spindle). But since most of that 5% failure is poor manufacture, out or on the edge of, tolerance, the HDD's per-year rate after that is precluded is quite a bit smaller.

Re:Poor statistics

[Anonymous Coward]

But regardless, my point was that "mechanical drives make a noise before they fail, therefore they're better than SSDs!" is a reeeeally stupid thing to say stands. Were you... disagreeing with this? Or did you just want to throw up your cred about managing a few hundred workstations because you're underappreciated at work? I can relate, by the way, but it's a bit off topic.

Fail. "[With most HDDs, it] will give you PLENTY of warning before it goes completely tits up" (the actual claim that you're trying to refute) is not the same as "mechanical drives make a noise before they fail, therefore they're better than SSDs!" (your deliberately-stupid straw man).

Re:Poor statistics

[Vanderhoth]

I think he's referring to the power supply in the machine, not just a right out all the lights in your house go off. I've had power supplies that just before they die altogether will flicker on and off repeatedly, that would cause the SSD to flicker as well, causing the issue outlined above.

Re:Poor statistics

[jcdr]

I have never see a warranty that protect the data in case of loss. The warranty only bring you a new device if the one you buy fail. You are lucky if the failed device is fixed instead of exchanged, as you might still found your data on it, but maybe not all coherent because of a failed transaction.

Re:Poor statistics

[Mashiki]

Well here's the problem, in some cases that extra money leads to higher performance throughput for various applications. So like any deployment it's a performance/cost ratio right? It might mean you're saving on X but degrading on Y, and still ending up having to deploy Z.

Re:Poor statistics

[Anonymous Coward]

Notably, his SSD failed, it didn't reach the write cycle cap.

Re:Poor statistics

[cdrudge]

But, for those of us that regularly recompile the OS and kernel, an SSD isn't going to stand up to that for very long.

Hogwash. There are many activities that write far more data on a regular basis to a SSD then compiling a kernel. Hang out in a HTPC forum and many, many people use SSD as storage for live tv buffers that are constantly writing a deleting GB of data every hour they are operating.

Re:Poor statistics

[Anonymous Coward]

This is precisely what I do as well. SSDs for the OS system and game volumes and RAID-5 magnetic disks for user and archival data.
Of course, this makes for a very heavy / large system to lug to a LAN party... but I don't do that much any more...

Re:Poor statistics

[Anonymous Coward]

Sure, it *seems* reasonable, but you're missing something.

When a manufacturer has offered a 5-year warranty for better than two decades, and then they suddenly *shorten* the warranty term, something is going on.

When it happens as they are struggling to get drive production back up after a devastating flood that virtually destroyed all their production capacity, you can be pretty sure what it is.

They've cut corners on quality control to get production numbers back up, meaning more drives are going to fail early. If quality control is cut severely enough, the 5-year warranty is going to end up costing them more money than their profits.

So now, instead of having drives that they're confident will make it 5 years in sufficient numbers to still profit after warranty claims, they're only confident that enough drives will make it 3 years to still profit after warranty claims.

Re:Poor statistics

[Anonymous Coward]

You started by making a point that is only valid when comparing drives using the same technology (and even then, only when using the same *generation* of that technology), and the moved on into supposition, then anecdote, and finally unsupportable statistical fantasy-land.

Nicely done.

Re:Poor statistics

[turp182]

Just to clarify, the hybrid drives will failover to the standard platters if the SSD portion fails?

Are there the same potential controller issues?

Just checking as I had a HHD fail in my laptop and I replaced it with a hydrid SSD/HHD (Seagate ST1000LM014 - a full TB in my laptop, and performance is very good).

Re:Poor statistics

[Miamicanes]

The POINT is that actual hardware failure of a SSD is unlikely, so they can offer long warranties, but despite their long nominal lives, they're more likely to fail early (and often, and repeatedly) at the one task you bought them for -- reliably and robustly storing data.

Re:Poor statistics

[h4rr4r]

If the drive fails, go get backups. Do not trust the data on a drive you believe has failed.

Re:Poor statistics

[h4rr4r]

Storage is free, IOPs cost money. Welcome to 2013.

Re:Poor statistics

[h4rr4r]

Really? Nagios will not notice all those machines offline? You do have monitoring servers watching the other monitoring services right?

How do you not notice for two days?
If any machine at any site does not respond in ~3 minutes I know.

Re:Poor statistics

[h4rr4r]

If you only care about storage space, take a look at $s per 1000 IOPs tell me how that works out. Failure rate per IOP would also be another bad story for HDDs since you need so many more of them.

Re:Poor statistics

[csnydermvpsoft]

Failing power supplies can do plenty of damage on their own, without triggering an obscure SSD failure mode. I had a power supply (a long time ago - I think it was an AT power supply [maybe even XT]) that took out a lot of other hardware as it failed; my guess is that it sent a voltage spike to the components.

Though, it's been at least a decade since I've seen a power supply fail. I try to put all of my machines on good UPS's - a refurbished APC SmartUPS is not too pricey.

Re:Poor statistics

[jcdr]

If I understand you correctly, there are two type of failure: a definitive one and a temporary one.
The definitive failure is very unlikely, so the longer warranty.
The temporary failure could raise early and repeatedly, but very difficult to prove.
This look like the bad sector problem on a HDD. As far as I know there is a similar mechanism used on SSD to mark bad region of a chip.
 

Re:Poor statistics

[Larry_Dillon]

Did they look at the length of warranty? Obviously it's not fair to compare a four year old spinning disk with five year warranty ot a one year old SSD with a one year warranty.

Re:Poor statistics

[Anonymous Coward]

And is his SSD still going strong ?

Re:Poor statistics

[nyctopterus]

Okay, but seriously, you should be backed up - relying on warning from your HDD is appalling stupid if you've got stuff you really need to keep. Weekly disk image? What is this, 2002? Seriously, good hourly incremental backups are really easy now, there's no excuse for this kind of cobbled together dodgy shit.

Re:Poor statistics

[mlts]

SSDs have their place. However, they need to have a good backup mechanism to go along with them. However, backups for all but the enterprise are still in the Stone Age. There are backup programs like Retrospect which have very useful features (deduplication), but it can be finicky about what optical drives it will work with (it will happily show you your new BD-R drive, but laugh in your face if you actually want to use it for anything.)

For UNIX, I'd probably give the nod to BRU just because unlike other backup programs, you can install them, and it doesn't demand a license key during the restore process (which can be a big catch 22 with other products unless someone has an "oh shit" binder in the tape safe with license keys in it.) For Windows, hard to say. The backup utility on the server side (wbadmin) is decent.

Of course, there is storing documents on a service like Dropbox and just rebuilding the OS and apps if there is a drive failure, but (and this is IMHO, of course), it is better to be able to restore the whole machine as an image than rebuild it from scratch, in a lot of cases.

Re:Poor statistics

[WD]

I tell folks if they want an SSD don't have anything on it they would feel bad if they lost
 
How about you tell people that it's unsafe to use a computer without a viable backup scheme, regardless of the type of drive they use?

Re:Poor statistics

[hedwards]

Where did I suggest that there aren't other things that use up a lot of write cycles?

What's more, those are applications that SSDs are not well suited. SSDs are fast, but HDD still have more durability in terms of write cycles.

Re:Poor statistics

[Anonymous Coward]

This raises a question in my mind...

(The following values are from Wikipedia.)

Most SSD's use a 2.5" laptop form factor. This size is (at maximum) 100mm x 69.85mm x 15mm = 104775mm^3.

Most HDD's use a 3.5" desktop form factor. This size is (at maximum) 146mm x 101.6mm x 25.4mm = 376773.44mm^3.

So the 3.5" desktop form factor is 3.596 times the volume of the 2.5" laptop form factor.

Since you don't need a motor or read arm/head, and since you don't need gaps between platters, that frees up a LOT of space that could be used for more flash memory. Controller board aside, the entire internal volume of the drive could be used for flash memory modules. Using a decent board-stacking arrangement with integrated heat dissipation to the metal frame of the drive, you could pack a LOT more flash memory into a desktop form factor, and probably relax some of the tolerances due to greater available space. They could save money on the frame itself, too. The frame no longer needs to resist vibration or torque from the motor or platters, and can be made of lighter, cheaper, and more heat-transferring materials.

This could both bring down the cost and raise the capacity of SSD's for desktop use, which would make them a bit closer to competitive with spinning disks in the $/GB ratio. It would also reduce the hassle of putting a 2.5" drive into a desktop that has brackets for 3.5" drives. The largest "consumer grade" SSD is 1TB. Wouldn't it be nice to have a 3.5TB SSD available, with higher capacities available as they make the chips smaller?

(For the record, there are some SSD's that come in a 3.5" form factor, but they're almost universally "hybrid" ones with spinning platters in the package with them.)

Re:Poor statistics

[Anonymous Coward]

I would rather have a more reliable drive with a shorter warranty. My RMA rate on SSDs is ridiculous, if they were not so freaking fast I would stop using them immediately.

Re:Poor statistics

[hedwards]

And if you have the code stored in an external repository, don't mind the small size and don't mind buying new drives regularly, then an SSD is just fine. For those of us that aren't rich or don't want to have several HDD, it's just not a reasonable solution.

Also the Linux Kernel is only about 30mb or so, I think it's under 100mb is you compile absolutely everything in. Compiling an entire OS is substantially larger in terms of final binary size and source code size.

Re:Poor statistics

[hedwards]

Spoken like somebody who doesn't understand the technology. That write cycle cap isn't so much a cap as a reflection of how many times you can use specific sectors. SSDs have hardware that reduces the likelihood of the same sectors being used continually. However, if you've got data stored on the disk, then those areas of the disk aren't available for wear leveling. So, the wear gets concentrated further. And thanks to the fact that SSDs are tiny compared with HDDs, you need to leave a much larger portion of the disk free to allow the wear leveling to work.

Consequently, you can have some of the drive no longer usable before the whole disk goes south.

Re:Poor statistics

[Anonymous Coward]

When SSD is virtually impossible to recovery any data. Usually with failed HDDs data can be recovered as the failure is result in loss of sectors that become unreadable.
NAND also degrades much faster than magnetic media. NAND gates can only handle a relatively low number of read-write cycles. Most NAND gates can only retain data for 2 to 5 years with out power, were HDDs can retain data for many decades (assuming the interface is available on newer hardware - as its nearly impossible to find an old RLL\ESDI controller on a modern PC)

Maybe when Resistive Memory SDDs become available SSDs will have a reliability advantage.

Re:Poor statistics

[Anonymous Coward]

Wrong.

Re:Poor statistics

[tlhIngan]

The other problem is the write cycle limit on the SSDs, that may or may not be an issue depending upon how you use your computer. But, for those of us that regularly recompile the OS and kernel, an SSD isn't going to stand up to that for very long.

You'd be surprised at the endurance of a lot of SSDs. Especially name brand ones like Intel and Samsung (forget the rest).

Some tests by users who do nothing but write to the drives 24/7:
http://www.xtremesystems.org/forums/showthread.php?271063-SSD-Write-Endurance-25nm-Vs-34nm

Like the old Intel 320s ,a 40GB one reported no more life at 190TB written, died at 685TB.
The Intel X25 80GB - worn out at 147TB, died at 883TB.

It's a challenge because they are damn hard to wear out so the newest and latest SSDs really take forever.

The SSD is likely going to be way too small for usefulness before one actually runs out of usable writes.

Re:Poor statistics

[Anonymous Coward]

Recompiling an OS doesn't represent a huge number of read/write cycles. Maybe 1.

Re:Poor statistics

[Full of shit]

Taking the hot/crazy scale onto account, what's the SSD analogue of "Fuck'em and dump'em"?

Re:Poor statistics

[hairyfeet]

Thanks for helping me illustrate my point in that you HAD PLENTY OF TIME to swap out the drive WITHOUT losing any data, with an SSD? It literally works one cycle and is dead the next, no errors, no warning, one of my gamer customer even had a health monitoring gadget in the hopes that it would be able to give him a heads up if the SSD was getting ready to die...nope, just went.

In a way it reminds me of the first years of HDDs, they too were prone to suddenly failing and would do so without warning but while HDDs got better every year thanks to all the shrinks if anything SSDs are getting worse. if you have all your data in the cloud or just don't mind the risk? Go for it, but until I see some better results with my own two peepers i'm gonna advise customers to wait.

Re:Poor statistics

[hairyfeet]

Make a clean box (cost around $30, Google for instructions) and place the platters from the drive with the data into an identical drive and it'll last long enough to get the data out. Try that with SSD chips and see how much useful data you get.

Re:Poor statistics

[larryjoe]

Having seen the actual warranty return rates on HDDs when I was working at an HDD company not too long ago, I can tell you that average ARR annual return rates (which are about 2x the actual AFR failure rates) may be somewhere around 5%, but that includes the worst as well as the best models. External drives are often floor-swept models that have higher failure rates, but the best models have ARR rates less than 1%, with actual failure rates less than that.

One other consideration for critical data is the variability in the time-to-failure distribution. With RAID configurations, low variability leads to an increased probability of data loss during rebuild. HDDs tend to have higher variability due to their mechanical nature.

Re:Poor statistics

[Zero__Kelvin]

"I would just add the whole thing ignores that big old rotting elephant in the room which is HDDs? I have found that in damned near every case, not all but most, will give you PLENTY of warning before it goes completely tits up whereas the SSD?"

I totally hear what you are saying. It is like in the transportation industry. It used to be that you could tell if your ride was going to fail, because your horse was looking sick. It was very rare for them to look healthy and then just drop dead in mid travel. These days you just never know when your ride might fail. I sure long for the good old days of reliable and predictable transportation failures.

Another great example is in the news industry. It used to be you could usually tell when your system was going to fail. You could actually see the letters getting fainter and fainter as time went on. Now all the letters on the screen stay the same from one day to the next and then one day ... poof ... there is a failure and you can't see any letters anymore. None ! I'll never understand why we just didn't stick with the old systems with reliable failure modes.

"Oh and one final thing....its gonna get worse"

I can only assume you mean this well happen after it stops getting better, as it has consistently been doing for some time now.

" its common knowledge that with each shrink the number of writes goes down and the number of failures go up and with all of the major chip companies seeming to only care about how many bits they can stuff per nano-meter?"

OH my God. You are being pretty selfish keeping this secret information to yourself! You should tell the SSD designers about this immediately! Maybe they can come up with some kind of scheme where they account for this; maybe some kind of system where they keep track of the writes, use error correction, and don't use the faulty cells!

"Call me old fashioned but I think I'll just pick upa caching SSD and keep the 5Tb in spinning rust, thanks ever so Intel."

Yeah. That's not really what I'm thinking of calling you right now actually.

Re:Poor statistics

[Zero__Kelvin]

You are forgetting one thing. None of this has anything to do with it. The OS could bee 100 Gigabytes, and it wouldn't make a damn bit of difference (excuse the pun.) It would take years of constant compilation to come anywhere near the limit on FLASH writes. Linus' SSD didn't fail due to too many writes, nor is that the typical failure mode:

25 nm, 80 GB SSD:

P/E Cycles: 3000
Total Terabytes Written (JEDEC formula): 68.5 TBW
Years till Write Exhaustion (10 GB/day, WA = 1.75): 18.7 years

Re:Poor statistics

[Zero__Kelvin]

" SSDs are fast, but HDD still have more durability in terms of write cycles."

Also, cars are fast but horses use less oil.

Re:Poor statistics

[Anonymous Coward]

And while scanning the SMART data is a nice start... you aren't going to get an e-mail when a branch office's first floor is under five feet of water.

I think you've missed the point. If they are under water then they all die. Both SSD and HDD. This says nothing about their reliability, and how well they handle failure due to use, wear, and manufacturing defects. And if the sites you work for have as many environmental damage events as you imply then please share who so I can avoid them like the plague.

Re:Poor statistics

[King_TJ]

As an I.T. guy myself who works on the side doing computer service, I agree completely.

I love SSD's as far as the performance goes, and basically swapped out all of my boot drives for them over a year ago, on my own machines. But I don't trust these statistics claiming they're more reliable than spinning hard drives. Like you say, there's more to data loss than just the fact that a storage device failed. SSD's just die suddenly, or as I observed in a few cases -- corrupt the system randomly and badly, every so many reboots. So you can go crazy doing reinstalls and data restores on a PC, thinking all is good again, until 3-4 weeks later -- boom, corrupt and non-bootable again!

Especially with more recent SATA hard drives and motherboards/BIOS's -- I've found the SMART technology often does what it's supposed to. You'll boot up and get a SMART failure message about a drive, even though the drive is still working. If you heed the warning and get your data off of it ASAP, you can just swap the drive before it ever really fails on you. Can't that that ever helped me with an SSD yet.

Re:Poor statistics

[Anonymous Coward]

I have several Segate drives that are 5 year also.

Are those cheap knockoffs of Seagate drives? You might have to worry about them being still around once you need that warranty then.

</sarcasm>

Re:Poor statistics

[WuphonsReach]

But, for those of us that regularly recompile the OS and kernel, an SSD isn't going to stand up to that for very long.

Mostly bullshit. Good steps to take are:

1. Use a file system that understands TRIM.

2. Get a good quality SSD. If you have a choice between the 128GB unit and the 120GB unit, go with the 120GB unit because it probably has more spare blocks in the first place.

3. Under-provision the device. If it's a 128GB unit, only partition out about 100GB of space. Leave the last part of the disk unused. Which gives the wear-level algorithm more "lightly used" blocks to play with.

There's currently three tiers of SSDs. The bargain-basement, do anything to get under $0.75/GB units. The middle-of-the-road units such as the Intel 330 which are aimed at the Consumer, but are not bottom-of-the-barrel designs. Then you have the enterprise SSDs which are designed for heavy write-loads and cost about $2-$3 per GB.

And that's ignoring the SLC SSDs, which are "industrial grade" and cost $5-$15 per GB.

Re:Poor statistics

[aliquis]

As far as what _I_ personally would care for what I would do would be to use the SSD as cache (possibly also as storage for the OS and the application I use the most at the moment) which would give much of the advantage of both for just about twice the price of either.

Re:Poor statistics

[WuphonsReach]

And while scanning the SMART data is a nice start... you aren't going to get an e-mail when a branch office's first floor is under five feet of water.

You will, if you have a proper set of environmental sensors hooked up to something like the Watchdog 15. They make temperature, humidity, power, water, intrusion sensors. The water sensor is a special cable that you lay on the floor somewhere (in a low spot) and which triggers an alert if it gets wet.

Re:Poor statistics

[WuphonsReach]

Enterprise grade SSDs have physical capacitors that store enough energy so they can flush any writes and such into flash memory when they lose power.

See tear-downs of the Intel DC3500 / DC3700 series drives.

Re:Poor statistics

[zwarte piet]

Yeah, I would like to know that too.

Re:Poor statistics

[kesuki]

i wanted to pick a few bones. first off there is this handy thing called S.M.A.R.T. (Self-Monitoring, Analysis and Reporting Technology) this technology works great in spinning metal but not as well for flash memory. now about the return rate, linus lost all his data But was able to Secure reformat the drive and it recovered.

spinning metal has it's flaws. but if the device can stop responding then work again after being reformatted suggests that the flaws in flash memory are individual sectors failing which on a spinning disk usually means backup then RMA, now it went from 'uh oh format lose data fail' to call tech support and get asked 'have you tried formatting it yet?'

Re:Poor statistics

[Anonymous Coward]

Since you don't need a motor or read arm/head, and since you don't need gaps between platters, that frees up a LOT of space that could be used for more flash memory. Controller board aside, the entire internal volume of the drive could be used for flash memory modules.

You're assuming that the internal volume of a 2.5" form factor storage device is a significant limiting factor on SSD size. That is a very bad assumption.

Using a decent board-stacking arrangement with integrated heat dissipation to the metal frame of the drive, you could pack a LOT more flash memory into a desktop form factor, and probably relax some of the tolerances due to greater available space.

You're also assuming that heat is a significant limiting factor. It isn't.

They could save money on the frame itself, too. The frame no longer needs to resist vibration or torque from the motor or platters, and can be made of lighter, cheaper, and more heat-transferring materials.

These things are already true of 2.5" form factor SSDs. The "frame" is usually a flimsy shell, far less robust than any 2.5" HDD frame.

This could both bring down the cost and raise the capacity of SSD's for desktop use, which would make them a bit closer to competitive with spinning disks in the $/GB ratio.

You're assuming that the frame and so on are significant cost drivers in a large SSD. They aren't. The cost of the flash chips is.

It would also reduce the hassle of putting a 2.5" drive into a desktop that has brackets for 3.5" drives.

Nobody really cares about this.

The largest "consumer grade" SSD is 1TB. Wouldn't it be nice to have a 3.5TB SSD available, with higher capacities available as they make the chips smaller?

You can easily fit multiple TB of flash in a 2.5" chassis. The reason it doesn't exist is that the price of flash memory is high enough that this would cost several times more than people are willing to pay for whole personal computers, so large SSDs tend to be for enterprise applications only, and in most cases people who want that much flash storage want it to be on a PCIe card so they can get higher performance than is possible through SATA.

(For the record, there are some SSD's that come in a 3.5" form factor, but they're almost universally "hybrid" ones with spinning platters in the package with them.)

Seagate is the only company manufacturing "hybrid" drives, and they're all (at present) 2.5" drives.

Please inform yourself, your opinions on this topic are crazy because you have no idea what you're talking about really.

Re:Poor statistics

[Anonymous Coward]

Spoken like somebody who doesn't understand the technology. That write cycle cap isn't so much a cap as a reflection of how many times you can use specific sectors. SSDs have hardware that reduces the likelihood of the same sectors being used continually. However, if you've got data stored on the disk, then those areas of the disk aren't available for wear leveling. So, the wear gets concentrated further. And thanks to the fact that SSDs are tiny compared with HDDs, you need to leave a much larger portion of the disk free to allow the wear leveling to work.

Consequently, you can have some of the drive no longer usable before the whole disk goes south.

Spoken like someone who doesn't understand the technology. That which guards against excessive wear isn't hardware, it's firmware. And SSD firmware can (and routinely does) shift static data around as needed to distribute write wear evenly across the entire media, even though this causes "write amplification" (which means that the drive is physically writing more data than the host computer actually requested).

The only "SSDs" which ever behave as you've described are simpler ones for simpler and less demanding applications, e.g. the "SSD controller" that is embedded inside of a SD card. (That plus the tendency to use much lower grades of flash is why it's not wise to boot from SD cards and the like if you value their longevity.)

Re:Poor statistics

[MikeBabcock]

There's nothing stopping an enterprise hard drive from doing the same thing -- my point was that its not necessary to add additional components on the spinning disc version. More importantly, my point was just that there are very few other benefits to spinning discs in the first place.

Re:Poor statistics

[Anonymous Coward]

2. Get a good quality SSD. If you have a choice between the 128GB unit and the 120GB unit, go with the 120GB unit because it probably has more spare blocks in the first place.

This one is yes and no. For example, consider the choice between a 120GB Samsung 840 and a 128GB Samsung 840 Pro. Does the 840 have more spare blocks? Yeah. But the reason for that choice is that the 840 uses 3-bit-per-cell flash memory, which has inherently reduced write cycle lifespan compared to the 2-bit-per-cell flash used in the 840 Pro. (It's also cheaper per bit of storage, which is why the 840 Pro is more expensive). Basically, the extra spare blocks in the plain 840 serve only to try to extend the life of inherently lower-life media, and the 840 Pro should still have a longer lifespan.

And also as you noted you can simply choose to under-provision a "128GB" drive to achieve the same effect as a competing "120GB" drive.

Re:Poor statistics

[smash]

Of course. Not disputing that. I am however disputing the article's premise that SSD is more reliable than hard drive storage. If you compare like for like in terms of the amount of data stored, SSD has a long way to go. If you compare like for like in terms of data stored plus cost, it's not even close. Of course performance is better. That is not in dispute.

Re:Poor statistics

[smash]

Whilst I agree IOPs are what counts - storage is not free. I gather you are referring to cloud provided storage, which has its own non-$ cost associated with it.

Re:Poor statistics

[hairyfeet]

You DO realize that what you are pointing out makes SSDs an even WORSE idea, yes? because it ISN'T the cap that is killing these things like flies, its the controller failures which have ZERO predictability so makes the entire discussion of the cap pointless.

At least with the HDDs you DO get advanced warning nearly every time, Windows delayed write failures, temp spikes, noise, something. With the SSDs the controller can crap any second, doesn't matter if its brand new or a year old and in fact one of my customers had one go tits up after less than a month, because there is serious problems with the controllers that nobody has figured out how to fix.

Oh and one final advantage of HDDs over SSDs, if you have a drive under warranty that starts going wonky? You can in most cases zero out the drive before sending it back, with an SSD failure I've had several customers just eat the cost and toss the warranty because there was no way to remove the data on the drive and they had no idea where or whom would end up with the drive once they sent it back. With an SSD you either use full disc encryption (which frankly slows the hell out of most drives, thus negating the whole point of having SSD over HDD) or you risk your data being recovered by whatever third world company refurbs the things.

Re:Poor statistics

[hairyfeet]

Yes and please post a reply to the guy below you saying the same so he knows as well. i have set up both hybrids and the Sandisk caching drive for Win 7 (if you have a desktop they are only $50 and a cheap way to get a boost WITHOUT having to reinstall or move the OS) and with both if the SSD portion fails? The original data is on the HDD and since it uses a technique similar to Windows Readyboost, which just FYI you can use instead of the sandisk caching software if you prefer, the data is never lost, it simply slows back down to HDD speed.

Re:Poor statistics

[hairyfeet]

I have gamer customers that "Must be teh bestest", one so bad his grandma uses a Skulltrail as that was the weakest hand me down he had, and I can honestly say...I have yet to see SMART work with an SSD, not even once, and we are talking dozens of dead SSDs.

They can call me names like Luddite all they want, I'll tell them the same as I tell win 8 apologists "You can take an innovative shit but its still gonna stink and you shouldn't play with it" because i can say with total honesty, and this is including the refurb drives that I used to get by the load off of Geeks (RIP Geeks.com, you are missed) that the ONLY HDDs I have seen in recent years that didn't give the user warning were the ones where the user killed it, for instance their kid knocking the laptop off the table. The rest? They had enough warnings shown to them to call me and have me replace the disk before data loss ever occurred. The reason why is obvious, the tech is mature, the bugs ironed out long ago.

Most people don't give a rat's ass about the drive, IT'S THE DATA and all it takes is having to tell one person that have their SSD go tits up before the next backup that their stuff was gone forever for you not to recommend SSDs.

Re:Poor statistics

[Zero__Kelvin]

Nothing makes an SSD a worse idea than HDDs. You either have never used an SSD or you have money invested in HDD stock. There is no other argument for using HDDs as your primary drive rather than an SSD in 2013. To say otherwise is absurd. The money saved in using SSDs far outweighs the cost of replacement in the unlikely event of a catastrophic failure.

Re:Poor statistics

[swalve]

Your first mistake was trusting any storage medium, or relying on the notion that you might get some warning when the drive fails.

Re:Poor statistics

[Reziac]

Which is all dandy cuz that's data you aren't going to need next year, or give a damn about backing up.

Re:Poor statistics

[hardwarefreak]

Enterprise grade SSDs have physical capacitors

Are you sure they're physical caps, not virtual, or ethereal caps?

Re:Poor statistics

[hairyfeet]

Are you stoned? WHAT money saved? The cost per Gb makes SSD much worse than HDDs, the controller failures (see "coding horror hot/crazy scale" for several PAGES going over 2 years listing failure after failure) makes them more expensive and if you data is worth ANYTHING you will have to use full disc encryption (thus negating a lot of that SSD speed) or just give up your warranty because while its pretty easy to zero out a HDD that is dying when an SSD controller fails there is NO WAY IN HELL to wipe the drive so I hope you don't mind some refurb factory in Taipei getting access to your SSD's data.

So I'm sorry but there is NO "savings" for using SSDs, like Ferrari its speed above all, above reliability (which will get worse as the shrinks and levels of MLC make corruption much easier), above capacity, in fact you honestly won't save any on power with a modern OS that isn't RAM starved as Win 7 and above can simply cache most used files in RAM and park the drive.

Re:Poor statistics

[hairyfeet]

Uhhh..as others pointed out I said it "will give you a warning" NOT "HDDs make noise you can hear" in point of fact you can get an SMS, or email, when a drive exceeds a limit that YOU set.

Try HDDLife Pro, last I check they have free trials so it won't cost you anything to try it and it will give you warnings if your drive starts having ANY SMART faults, not just low enough to trip a SMART warning, and will also warn you if the drive starts climbing past its normal operating levels. I use it with my SMB customers, works great.

Oh one final thing...I've dealt with over 20 failed SSDs and not a single one, not one mind you, threw any SMART errors before failure. Instead you would literally find it worked one day, dead the next. With the HDDs I've probably had who knows how many over the years with various customers die and in the past decade I have yet to come across a failing HDD that was impossible to get the data off of, if no other way you could always throw together a clean box and change the platters. i have yet to hear of anybody recovering data off an SSD without paying a top notch recovery company to the tune of a couple of thousand dollars.

Re:Poor statistics

[Zero__Kelvin]

You are so far off base it is ridiculous. You clearly don't do anything remotely useful with your computer hardware. Personally I improve my productivity drastically by not waiting for the slowest part of an antiquated computer system, to wit, the rotating platter based hard disk. I know you sell toys to people who play games. This is a discussion for the big boys who make a living with their systems. For us, a 20% increase in productivity translates to $20,000+ per year savings, and 20% is an exceedingly conservative estimate on the productivity improvement realized by professionals in the industry. Furthermore, you don't seem to understand the difference between 1% and 5%. Your claim that SSDs are less reliable than HDDs is directly contradicted by the facts, and your claim this it is going to get worse instead of better, as it has been for some time, is ludicrous. I find it especially ironic that you admit you don't save any money using SSDs (in other words, they are toys to you) and then go on to complain that you just have to encrypt your data. Are you really that afraid that someone will get at your WoW data?

Re:Poor statistics

[h4rr4r]

No, I mean if you are buying disks you get lots of storage in 2013 no matter what you select. The IOPS however are very expensive.

Re: Poor statistics

[RockDoctor]

Yep, if in the slightest doubt, stick a UPS in there and let that die before the important stuff does.

I was running an experimental computer (OS 9000? Long time ago, and this tablet isn't good for flipping tabs to research stuff) years ago on an industrial site with an under-powered, over-loaded desisel genset. When the big machines switched on, the fluorescent lights would flicker and die ; 250V nominal AC dropped to under 100V, and the supply was absolutely filthy with noise from the SCRs. We used a low-pass filter to try to take out the high spikes, and a high-pass filter to soften some of the sags, feeding a UPS, feeding a second different type of UPS, feeding power to the computer.

We burned out 3 UPS devices in under a week (and replaced each one; fortunately we were on shore not at sea), and then the computer's power supply burned and we went back to pencil and paper.

I suggested just getting our own generator, but that fell foul of the flammable atmosphere regulations.

Re:Poor statistics

[alexo]

Hmm, my drives send me emails when they start having problems.

You're lucky. My kids' drives just update their FB status while my parents' drives calls at 2am complaining that I should visit more often.

Re:Poor statistics

[Decker-Mage]

Sorry, the population count doesn't change the statistic for large population numbers, say greater than a thousand or so. And even then, you're given a very small error band, plus or minus. That's (extremely) basic statistics. [Population vs. T-test in case you don't know.]

Re:Poor statistics

[stoatwblr]

Big, Fast, Cheap. Pick any 2 - that applies as much for spinners as SSDs (enterprise HDDs are generally almost as expensive as SSD)

If you're willing to compromise then solutions like ZFS help a lot.

Bear in mind that raid is no substitute and vice versa. (The best backup being the one you never need to use, but having to restore backups is painful - in our enterprise-scale academic setup 99% of restores are for material that was deleted over a week ago and is suddenly needed again.)

Re:Poor statistics

[stoatwblr]

"I may be just a little country shop guy but when my gamer customers have all experienced multiple failures when it comes to SSDs, and these guys don't go cheap"

Don't kid yourself about gamers. They're bleeding edgers for the most part and there's an old saying that "pioneers get arrows in their backs"

Having said that I've seen at least one consumer SSD which would shut down (*poof*) after a couple of hundred hours powered up and then stay dead if tried again. Leave it unpowered for a few days and the story changes: Power it up and it's fine for a while (subsequent uptime is proportional to a multiple of the powered down time) with no actual data loss.

This symptom appears to be more widespread than manufacturers realise (I've seen a lot of reports of this kind of behaviour across a wide number of makers) - and of course by the time such a dead drive gets into their hands it looks and runs ok, so it gets tagged "no fault found" and returned to the consumer (who finds the fault's not fixed and chucks it in a drawer vowing to never buy from that maker again)

Do the math

[djupedal]

5% of a 4TB HDD that sells for USD$200 is roughly 200 GB = $10.

1.5% of a 4TB HDD that sells for USD$29,000 is roughly 60 GB = $425.


I can guess who is pushing these casual comparisons, but seriously - when price parity kicks in, let me know. In the mean time and as a developer, I have no use for SSD in my desktop system.

Re:Do the math

[Anonymous Coward]

Alright, I'll do the math....

9ms average access times on a 7200RPM spinning drive == ~100 IOPS.
High-end SSD: 100K IOPS.

Yes, a thousand times the number of disk accesses. If you're really a developer, you'll see your compile times cut by a factor of 5-10 (depending on how much CPU power you have to spare). Things load from disk like magic.

You don't buy SSDs for the raw capacity, you buy them for the *fast* access times. Period.

Re:Do the math

[dunng808]

Is 4TB representative? Or are you just putting more spin on this story?

Re:Do the math

[Desler]

9ms average access times on a 7200RPM spinning drive == ~100 IOPS.
High-end SSD: 100K IOPS.

The SSD that most consumers are using are neither high end nor have such IOPS ratings.

Re:Do the math

[Anonymous Coward]

> as a developer, I have no use for SSD in my desktop system.

Do you compile code?

Re:Do the math

[camperdave]

> as a developer, I have no use for SSD in my desktop system.

Do you compile code?

SSDs are for booting. RAM disks are for compiling, and hdd is for long term storage.

Re:Do the math

[Anonymous Coward]

*shrug* As a developer, installing an SSD paid for itself in time saved due to waiting on disk within three months.

Seriously, git is instantaneous, greps and compiles are ludicrously fast, etc,, etc, etc. I mean, unless your rate is like 10USD/hr, you owe it to yourself and your clients to install a decently-sized decently-fast SSD in all of your dev boxes.

Re:Do the math

[mysidia]

5% of a 4TB HDD that sells for USD$200 is roughly 200 GB = $10.
1.5% of a 4TB HDD that sells for USD$29,000 is roughly 60 GB = $425.

You mean 5% of the space size in GB is that.

Your math is wrong, because you are misinterpreting the statistics. A 1.5% SSD failure rate, with a small number of disks, does not mean that "1.5% of the capacity" fails; if you purchased N 4TB SSD "that sells for 29k"; on average N*1.5% Of those entire SSD drives fail; and if you purchased N 4 TB HDDs that sell for $200, N*5% of those entire HDDs fail.

Due to I/O constraints; when you use HDDs, you don't get to use all your total space, before performance degrades to unacceptable levels, and you have to buy more HDDs; furthermore, all those extra HDDs consume a lot more power than SSDs. The $/IOP is not attractive for HDDs: the vast majority of computer users do not need 4TB HDDs; and will use 100 to 150GB TOPS.

Therefore: SSDs look a lot more attractive, when you discount, or forget the existence of the portion of the capacity that the user cannot use due to performance constraints, or will not use -- due to not needing the space.

Last I checked; you cannot go to Amazon, Newegg or your local supermarket and buy a 200GB hard drive for $10 It is not an option; the least cost new HDD you can pick up is about 60 bucks. However, you definitely can go to Newegg, and buy a new 150GB SSD for about $250.

Also, the Crucial M500 1TB SSD is $600. 4 times that is $2400, not $30,000.

4TBs are for archival purposes, where the hard drive will be powered off most of the time, anyways. The failure rate of 3 TB and 4 TB HDDs is probably much higher than the 5% average, due to the tighter mechnical tolerances and higher density encoding methods required. I believe the 5% figure applies to 1.5TB disks.

Re:Do the math

[gman003]

Actually, you'd be surprised. The Samsung 840 EVO, a low-cost consumer drive (the high-end is the 840 Pro) that gets down to $0.70/GB, can hit 90K IOPS read on every model, and 90K IOPS write on 500GB models and up.

Sure, older or ultra-cheap drives won't hit that (my new Chronos doesn't get there), but rounding to the nearest order of magnitude will get you 100K IOPS even on medium-end consumer drives.

Re:Do the math

[10101001 10101001]

More to the point, you can buy 4 4TB HDDs for $800 and setup a RAID1 and get a lot of the same read performance as an SDD while having heavy redundancy. Yes, you don't get the same level of gains for writes nor do a lot of systems support 4HDs readily (not enough physical space to place them) and there's the issue of power consumption. But if there's a 5% of failure with one HD, then presuming there's no relationship to the drives (ie, you don't just buy a bunch of drives from the same company which were all from the same batch production), then you'd expect a very low probability that all the drives would fail (something like 0.000625%). Besides that, you could very trivially buy multiple extra HDs for backups and still be way cheaper all around.

Re:Do the math

[brit74]

A while back Joel Spolsky (joel on software) tried switching to an SSD and compared his compile times to his old HDD. The result: no difference. Apparently, the disk access isn't the slow part of the compilation process. The bottleneck in compiling seems to be the processor speed.

Re:Do the math

[Anonymous Coward]

I guess you won't be buying any RAM for your desktop system, either.

Because that costs more per byte, too.

Sure, it's faster - but it costs more. And I guess that's all that matters.

Re:Do the math

[dbIII]

After "doing the math" a mixture works - cheap small SSD for system files and a pile of spinning disks with ZFS to hold stuff. Of course on MS Windows with the C: drive nonsense still hanging about a mixed system is a bit more difficult and may mean a bigger and more expensive SSD for those programs that will only work on the system drive.

Re:Do the math

[Rockoon]

You might want to factor in that SSD's often have longer warranties than HDD's these days.

OCZ's SSD's are 3-year while Intel SSD's are 5-year. HDD's manufacturers reduced their warranties from 3,4, or 5 year to 1, 2, or 3 year in 2011.

I'm not saying that thats the situation of the data in the study, but it could be. 5% on an average 2 year warranty vs 1.5% on an average 4 year warranty, well that is quite a significant difference.

Re:Do the math

[c0lo]

> as a developer, I have no use for SSD in my desktop system.

Do you compile code?

With a 32 GB RAM, why do you thing compilation would be relevant for the SSD or HDD choice?

Re:Do the math

[c0lo]

Or... you know.... actually put some RAM into your developer machine.

Re:Do the math

[Anonymous Coward]

HIGHLY HIGHLY dependent on the compiler and the language. C/C++/Fortran and many other traditional compiled languages generally can be compiled so fast that disk speed is VERY MUCH a factor.

With Java and C#, RAM is much more the limiting factor. I'm sure it ALL depends on the specifics of what you are doing.

Re:Do the math

[wolrahnaes]

Everyone's already jumped on you on this one, but I still have to chime in and say holy crap are you ever wrong if you think a SSD is not worth every dime right now.

You don't need *everything* to be on SSD, just the commonly-used data; your system, binaries, etc. I have a 120GB Intel 330 series SSD (though in hindsight I'd rather have gone with something ~250GB and 2TB of normal spinning disk, plus a 7TB server. The system, my home directory (minus downloads directory), and my main games are on the SSD, the rest of my games are on a 1.5TB 7200 RPM disk, and downloads get an old 500GB disk. My machine spends more time in POST than the entire rest of the boot process, browsers launch instantly, IDEs take half a second, games practically skip their own loading screens, etc.

Unless you place no value on your time, in any interactively-operated machine where you can have multiple storage drives a SSD is a no-brainer. Laptops where a second disk is challenging or impossible it entirely depends on your mobile storage needs, but the reduction of moving parts and performance/battery improvements are a strong argument in favor of them.

Re:Do the math

[Anonymous Coward]

200 IOPS a drive on average for SAS drives. SATA are often closer to 100-120 IOPS a drive != 90k IOPS. SSDs are crazy fast when it comes to small random IO.

Re:Do the math

[MikeBabcock]

Even the 840 Pro isn't that expensive actually, just ordered one today.

http://www.newegg.com/Product/Product.aspx?Item=N82E16820147193

Re:Do the math

[Dahamma]

Wait, you are basing the improvements in compile times on one guy's anecdotal results? Well, here's another: when I switched to an SSD at work my compile times were cut by more than half. It was an huge difference in compile time ie. productivity.

It all depends on your codebase and tools, really. He was probably compiling a relatively small codebase, and for all we know his methodology sucked so a lot of it was in the RAM cache. I can tell you for a fact that a clean build on a large code base was drastically improved.

Re:Do the math

[MikeBabcock]

For any reasonable code size, the data's all cached in memory already from you editing and saving the files in question.

Re:Do the math

[MikeBabcock]

No he's doing the math right -- At an annual failure rate of 1%, you need to replace 1% of your total capacity every year. With an annualized failure rate of 5%, you need to replace 5% of that capacity overall. The averaging is done because over time, it works out, just like insurance. Sure, on any given year *if* a drive fails, you have to pay for the whole thing, but that's not how one accounts for such failures.

Re:Do the math

[Mad Merlin]

More to the point, you can buy 4 4TB HDDs for $800 and setup a RAID1 and get a lot of the same read performance as an SDD while having heavy redundancy.

Where by "a lot of", you mean less than 1% of, right?

Typical IOPS on a 7200 RPM HDD is around 80. Typical IOPS on a garden variety SSD is 80,000. We'll be generous and assume linear speedup for the four HDDs, which gives us 320 IOPS, or 0.4% of the performance of a single SSD.

Re:Do the math

[AcidPenguin9873]

How often does a developer need to do a clean build? If your build flow is set up right, not very often. Especially not often when the developer is in the critical code/compile/fix-compile-errors or compile/run/debug/fix loops which are where the compile times are important. There are even build flows/tools that let you not ever do a clean build, by storing versioned object files. And if your codebase is large enough, hopefully you're using dynamic linking which means that the link step isn't too long either.

Re:Do the math

[Dahamma]

Why on earth do you need 4TB "as a developer"?

And it's absurd to claim that SSDs aren't useful over HDDs just because HDDs are more cost effective by space. HDDs are also more cost effective by space than RAM - so why not just minimize your RAM and create a huge swap space?

Re:Do the math

[GigaplexNZ]

I upgraded to an SSD recently and compile times didn't really change at all, which was surprising. Boot times, application load times, system responsiveness etc however are massively improved. At least on my setup (Xeon E31270, Windows 7, Visual Studio), compile times aren't particularly IO limited.

Re:Do the math

[Dorianny]

According to Research from Segate, a hybrid drive needs just 8gb of NAND to achive %95 performance of a a NAND drive in a typical business environment "During the five days of study, the average amount of data read by machines in a business environment stood at 19.48GB. Out of this amount, just 9.59GB was unique; the rest consisted of duplicate reads" Of course this is not exactly a large scale study, but it was presented in a industry workshop so its not just fabricated marketing material either. http://www.techweekeurope.co.uk/news/seagate-hybrid-drives-dont-need-more-than-8gb-of-nand-124069

Since benchmarking software is useless for gauging the effect of caching, It would be interesting to see a similar study done on typical usage scenarios for a home machine.

Re:Do the math

[hobarrera]

Yet, on a laptop with a 128GB SDD, I've got ~95GB free. Why would I want to buy a huge HDD, if I've got more than enough with a tiny SDD?
Differente people have different needs, and, the point is, for those who want reliability for a brand new disk, SSDs are the way to go.

Re:Do the math

[GigaplexNZ]

Fairly large codebase here, ~4 minute compile times, C++ with Visual Studio. Compile times were unaffected by the SSD upgrade. Searching code, however, massive speed improvement and paid itself off with productivity improvements after about a month.

Re:Do the math

[Dahamma]

And everything you said reinforces my statement, "It all depends on your codebase and tools, really" :) Not everyone (in fact, probably very few) get to pick all of the tools and libs that they use...

I had to do a clean build today, actually. And it was unavoidable. I upgraded my PS4 SDK, and not doing a clean build when your entire libc/SDK/etc changes is practically a guarantee of random impossible to track down errors in your app. Due to quirks of the existing system, that particular build was almost 5x faster on my machine vs. a coworker's using a slower, non-SSD machine. 1/2 hour saved times 2-3 clean builds easily pays for the disk already.

Re:Do the math

[Anonymous Coward]

"Yes, you don't get the same level of gains for writes..."

Nor do you get the level of gains for reads that you claim to.

"But if there's a 5% of failure with one HD ... you'd expect a very low probability that all the drives would fail (something like 0.000625%)."

But there's four times the probability of drive failure over a single drive. Furthermore, you only need 2 drives to fail with RAID 1 to risk data loss, not "all the drives".

"Besides that, you could very trivially buy multiple extra HDs for backups and still be way cheaper all around."

It wouldn't be cheaper in terms of time. maintainance, size, power, or thermals. The performance would almost certainly be worse and that has a potential cost as well.

Re:Do the math

[GigaplexNZ]

I've been running multiple partitions/drives on Windows for years and haven't encountered a sizeable app that I'd want to put on a different drive that wouldn't let me.

Re:Do the math

[Anonymous Coward]

4TB of SSD is under $4000 today, not $29,000. At the higher end of the consumer range, barely under $4k; at the lowest cost/gb end of the range, under $3k. To spend $29k on 4TB of SSD, you'd be talking lower-end enterprise SAS SSDs; enterprise SATA SSDs are under $1000 for 400gb if you shop around, or about $10k for 4TB.

(Your HDD price is also too high for consumer drives, too low for enterprise drives.)

Re:Do the math

[blackraven14250]

There's one thing I'd like to mention about SSDs that you don't take into account. You can't use all of the space on an SSD either without performance degradation. I'm a SSD user myself, but it's just as true of SSDs as HDDs that you don't want to actually fill it past a certain point.

Re:Do the math

[10101001 10101001]

Typical IOPS on a 7200 RPM HDD is around 80. Typical IOPS on a garden variety SSD is 80,000. We'll be generous and assume linear speedup for the four HDDs, which gives us 320 IOPS, or 0.4% of the performance of a single SSD.

Great. Thanks for, like so many other people in this thread, pulling out a near meaningless benchmark figure*. As others point out, you might see 1000x the IOPS but real world results are more on the order of at most 5-10x speedup (a look at some of Tom's Hardwares HDD vs SDD seems to confirm it with adding music to WMP on HDD and on SDD and Gaming on HDD and on SDD). Why is that? Obviously because most applications and activities don't involve randomly accessing 4K files/sectors scattered all over the storage device. File systems are heavily designed to avoid fragmentation and most file accesses are linear and of considerable enough size that actual streaming performance is more critical. Hell, ReiserFS (both in synethic and real benchmarks) has shown that just having the file system actually intelligently deal with smaller files (NTFS does this with the MFT, AFAIK) nets most of the same benefits.

So the other point would be something about whether four HDDs would see a linear performance increase and see the 3x increase and start to approach SDD territory. Honestly, I actually doubt it. And I'm certain in some circumstances SDD would still heavily blow away what even a large stack of HDDs in RAID1 could do. But in most real world circumstances, the difference would be pretty damn negligible, except probably noticing how slow big file copying can be.

*Yea, I know this isn't an actually meaningless figure. It gives you some idea of just how much better random access to files will be. And if you have a specific workload that deals with lots of non-cached random files or sectors in large files, I can certainly see some clear benefits to SDD. But, honestly, coupled with things like caching, lots of RAM in system now days, things like Superfetch, and having just limits on just much data can actually stream over SATA, SDD isn't worth it to me or a lot of people--but then most people aren't going to do RAID at all and SDDs main benefit would be the no-moving-parts and lower power usage for portables. If you're one of the exceptions and can afford the massive extra cost, good for you.

Re:Do the math

[chuckinator]

RAM disks are for initrds and initramfs images.

Re:Do the math

[symbolset]

Research from Seagate says their hybrid drive only needs tiny ssd? Imagine that.

Re:Do the math

[Anonymous Coward]

Not really surprising, after the first run. Your source and up-to-date object files would mostly be cached.

Re:Do the math

[Anonymous Coward]

C++ compiles... fast? Did I suddenly get teleported to an alternate universe?

Re:Do the math

[Anonymous Coward]

It doesn't cost that much to get a relatively high performance drive of modest size. I've got a half terabyte one for close to $300 on a newegg sale. And the 128's can be had for around a hundred.

Re:Do the math

[fnj]

RAM disks are for compiling for those who don't really understand how a modern OS buffercache works.

Spectacularly poorly, in my experience, for the kinds of things I do.

Re:Do the math

[fnj]

C compiles very fast on modern computers. Fortran, I haven't any modern experience. C++ is spectacularly slow to compile if the code being compiled is at all sophisticated. And it's all compute. The drive light barely lights up for minutes at a time.

Re:Do the math

[BitZtream]

Then you're a shitty developer.

As a developer, switching to an SSD was the biggest improvement I've ever made to my workflow. My 2009 machine with an SSD out performs my 2013 i7 machine with a 6 disk raid 1+0 machine with 16GB of ram. (which now has an SSD in it as well)

The time savings of an SSD on source files and compiling is ridiculous.

My time is worth FAR more than an SSD, sorry yours isn't.

Re:Do the math

[gagol]

No, you are still in a universe in which you compile on a 386...

Re:Do the math

[BitZtream]

...

Compiling is almost ALWAYS IO bound for a project of any size. Why do you think make -j 4 or make -j 8 makes a noticeable difference on even single core machines? Because the compiler spends most of its time waiting on disk IO, reading and writing all those intermediate files.

VisualStudio also supports parallel compilers for this exact same reason.

Re:Do the math

[Electricity Likes Me]

And the thing to remember about all storage is that it will fail. If you have a single disk in a machine, and that machine is not backed up properly, then you will lose that data in the next 5 years.

Re:Do the math

[fnj]

It's hopeless. The rubes are mod'ing up GP, while parent is the one who is correct.

Re:Do the math

[snero3]

SSDs are for booting. RAM disks are for compiling, and hdd is for long term storage.

RAM disks are for compiling? how small are you projects?

Re:Do the math

[Anonymous Coward]

Nope. Still not right. The failure rate isn't tied to capacity, so the comparison is invalid. If a yacht has a 1% hull failure rate and a ferry has a 5% failure rate, the fact that the ferry can take on a lot more water before it sinks doesn't change a thing. 5% of them take on enough water to sink.

The annual failure rate reflects the entire volume of drives under warranty. That is, x% of drives fail each year. As certain units fall out of warranty, new ones are sold, but if the rate remains x% per year, then each year you've got a x% chance of failure.

That doesn't mean 3 years at 1% annual failure rate equates to a 3% chance of failure, nor does it mean that you can simplify it to "you need to replace 1% of your capacity every year".

Sure, on any given year *if* a drive fails, you have to pay for the whole thing, but that's not how one accounts for such failures.

Since it's not possible to replace 1% of your capacity if you only own a small number of SSDs, it's exactly how one accounts for such failures.

The GP's "price per GB" doesn't provide any useful information to anyone and doesn't reflect any sane form of risk modeling. The only possible point he's making is one that no one disputes: replacing an SSD is more expensive than replacing a hard drive.

If you've got one hard drive and one SSD, then this suggests the hard drive is more likely to die in 3-5 years than the SSD is. Nothing more, nothing less.

Re:Do the math

[rolfwind]

1.5% of a 4TB HDD that sells for USD$29,000 is roughly 60 GB = $425.

Idk about 4TB, but you can get a 1TB system for $635.
http://www.newegg.com/Product/Product.aspx?Item=N82E16820147251

In the mean time and as a developer, I have no use for SSD in my desktop system.

Dude, you don't know what you're missing. I use both, HDDs are now my data drives for archives, and SSDs are my primary boot drives and the data I'm immediately working on. I'm glad I'm working on HDDs anymore.

Re:Do the math

[dbIII]

Run a few dozen MS Windows systems with a variety of software instead of a single machine and you'll start finding them :( The biggest culprit in my case comes from Halliburton of all places and expects to put it's data files on C: as well, which is a pain when projects are a few hundred GB each. When you copy a project to another machine it has to be on exactly the same path, so big system drives are needed all round.

Re:Do the math

[fnj]

Even on Windows it's easy to move "My Documents" off of C:. "Program Files" is not the big barrier in most cases. FAT32 is long dead for system drives, and NTFS has symbolic links.

Re:Do the math

[DigiShaman]

$0.70/GB is very good for mid-range SSDs. 1.00/GB for top-range SSDs is affordable now, but the laws of diminishing returns still applies as it does with all other forms of newer technologies.

Re:Do the math

[0123456]

Because the compiler spends most of its time waiting on disk IO, reading and writing all those intermediate files.

That's why smart people put any intermediate files in a RAM disk.

Re:Do the math

[fnj]

Why on earth do you need 4TB "as a developer"?

Depends on what they're developing, now, doesn't it? Maybe a huge database. Maybe video processing code. Heck, I'm not even primarily a developer and I currently have 19TB on my primary desktop and 32TB on two servers.

Re:Do the math

[0123456]

RAM disks are for compiling? how small are you projects?

How big are yours, if the generated files won't fit in a few gigabytes of RAM?

Re:Do the math

[Yosho]

Why do you think make -j 4 or make -j 8 makes a noticeable difference on even single core machines?

Uh... in my experience, they don't make much of a difference. They slightly increase performance due to one process being able to compile while another process is doing disk I/O, but it's not a huge difference. Shouldn't your build system be piping data between processes rather than writing intermediate files out to disk?

The only step of the build process on my C++ projects that is significantly I/O bound is the linking step, and even then, it's only an issue if I'm linking in a large static library.

Re:Do the math

[k8to]

Sure, you can compile in RAM, (my commercial work project with a ~4GB build tree does this whithout a ramdisk just fine) but linking in ram is right out.

And since in C++ linking is the slowest part by FAR on rotating disks, SSD offers an immense benefit.

If in some other language like pure C or ocaml, then this may be less true.

But then again just building the dependency graph for make wins so hard on SSD, there's still no reason to forgo it. And if you have to build ANYTHING with visual studio, the whole-project link-time optimizations make linking so expensive that you really desperately want the SSD.

Re:Do the math

[epyT-R]

Depends on many things, including the compiler, the language, the relative size of the project and how many files it's spread over. It can be somewhat limited by both..ie an increase in performance of either storage IO and/or CPU/ram offer boosts.

Re:Do the math

[0123456]

Even on Windows it's easy to move "My Documents" off of C:.

Uh, no, it's not.

Kludgily moving some files out of there to a different drive is easy enough, but getting the whole user tree onto another drive is a huge pain; I had to boot into some special hidden mode during the Windows install to do it. And some programs still continue to write to the C: drive regardless.

Re:Do the math

[Yosho]

Having a lot of RAM isn't going to make the initial read of files from disk or writing to them any faster.

Re:Do the math

[WaffleMonster]

Seriously, git is instantaneous, greps and compiles are ludicrously fast, etc,, etc, etc. I mean, unless your rate is like 10USD/hr, you owe it to yourself and your clients to install a decently-sized decently-fast SSD in all of your dev boxes

Since when is compilation an I/O limited activity?

Re:Do the math

[0123456]

And since in C++ linking is the slowest part by FAR on rotating disks, SSD offers an immense benefit.

How?

You've just built those files, so they're sitting in the disk cache. Is writing out the linked file really that slow?

Oh, hang on, you're compiling in Windows. Doing a clean build of our C++ source tree in Linux, we spend about 99% compiling and 1% linking.

Re:Do the math

[hawguy]

5% of a 4TB HDD that sells for USD$200 is roughly 200 GB = $10.

1.5% of a 4TB HDD that sells for USD$29,000 is roughly 60 GB = $425.

I can guess who is pushing these casual comparisons, but seriously - when price parity kicks in, let me know. In the mean time and as a developer, I have no use for SSD in my desktop system.

If capacity is all you're looking for, SSD lags spinning media. But few people buy an SSD for the storage capacity, they buy for the IOPS.

A $300 15K rpm Cheetah 450GB SAS drive will give around 200 - 300 IOPS.
A $400 Seagate "600" 480GB enterprise SSD will give around 11,000 - 80,000 IOPS

So, 50 - 250 times better performance for 33% greater price. I'd say that price parity is already here.

I'm surprised that as a developer that you see no use for an SSD in a desktop system. I could understand not putting on in a laptop where you might have only a single drive bay, but in a desktop there's usually space to tuck one in somewhere even if there's no free drive bay. And the performance gains are quite noticeable.

Re:Do the math

[B1oodAnge1]

Sounds like using symbolic links would help you out.

I've been storing my ( >600 ) steamapps folder on another HDD for years.

Re:Do the math

[Anonymous Coward]

If you keep your build machine running a clean build would also be in RAM cache.
With enough ram all your source code would get into cache the first time you compiled something.
Even when you clean build all your source code is still in cache, all the object code is written into cache first.

SSD for development will not matter at all, at least if you use a modern operating system.

Re:Do the math

[AcidPenguin9873]

When you had to do this clean build, were you in a critical code-writing/code-debugging loop? That was my main point. How often do you upgrade all the libraries in your project necessitating a clean build AND you're in a critical build loop?

Re:Do the math

[Dahamma]

Yes, but that has NOTHING to do with using an SSD to improve performance, unless you have 19TB on your primary desktop with a single HDD, which obviously you don't.

Re:Do the math

[drkstr1]

I can tell you have never used an SSD. At a certain point, more RAM becomes hardly noticeable. That SSD drive I got was the single most noticeable improvement in productivity I've gotten out of a computer part.

Re:Do the math

[GigaplexNZ]

Compiling is almost ALWAYS IO bound for a project of any size. Why do you think make -j 4 or make -j 8 makes a noticeable difference on even single core machines? Because the compiler spends most of its time waiting on disk IO, reading and writing all those intermediate files.

From my experience when I toyed around with Gentoo a few years ago, make -j worked best when the number provided was {core count + 1}. Any more than that and it slowed down due to context switching overhead and/or disk head thrashing.

VisualStudio also supports parallel compilers for this exact same reason.

It has severe limitations. It doesn't work when you need to import TLB files, it does not work with incremental rebuilds, and it doesn't play particularly nicely with precompiled headers. The former is required for our project to build at all, and the latter 2 improve productivity more than using multiple threads (YMMV).

Re:Do the math

[Dahamma]

It doesn't matter. Time is money, and SSDs are freaking cheap compared to developer time. While I was not completely idle the whole time, I was less productive since I was mostly concerned with a new version that supposedly fixed some critical issues. And even with an incremental build 20-30 seconds 30+ times a day over a large project adds up!

Honestly I have no idea what your point is here. Are you honestly trying to tell me with almost no information into my situation that an SSD is not helping *my* productivity, or are you just bored and trying to be contrarian?

And again to reinforce the "It all depends on your codebase and tools, really" - dynamic linking is most definitely NOT relevant to all projects. Sometimes it's worth the performance gain not to do it (especially with embedded systems or game consoles) and sometimes it's not even possible with the architecture....

Re:Do the math

[grumpyman]

And tape is for infinite amount of porn?

Re:Do the math

[Anonymous Coward]

lie to it.

use the mklink command in windows

create a hardlink from c:\whatever\wherever to anywhere you want. and the pc won't notice or care the path doesnt really point to c:

Re:Do the math

[c0lo]

I can tell you have never used an SSD. At a certain point, more RAM becomes hardly noticeable.

And for a heavily C++ templated app, the IO is almost irrelevant (RAM and CPU predomine).
While for a Java project, Eclipse compiles the source while saving (me pressing Ctrl-Save is likely to take more than the compilation and writing the file to disk).
Not to mention that, for many compilations, the intermediary files may safely be generated on a RAM disk.

Yes, sure you are right: didn't use one because I never felt the need for an SSD while dev-ing. Certainly, the fact I'm not required to use Visual Studio or the like helps.

Re:Do the math

[ranton]

If your compile time is 8ms, you are probably compiling some school project.

Re:Do the math

[Andtalath]

The 840 evo is a piece of crap, it has got very few write cycles.

Re:Do the math

[VortexCortex]

RAM disks are for compiling for those who don't really understand how a modern OS buffercache works.

RAM drives (not disks) are for compiling for those who understand the difference between circles and squares.
Modern OS IO caching is a crutch for poorly written software systems that should be using a daemon if they need that performance.
Additionally: Firmware is for /boot/ for those who want instant-on and the same security of UEFI (even on x86) without the cluster-fsck of UEFI.

Re:Do the math

[nedlohs]

Since cpus have been so much faster than disks.

Re:Do the math

[Anonymous Coward]

With a 32 GB RAM, why do you thing compilation would be relevant for the SSD or HDD choice?

GCC will eat 32GB of RAM quite happily thank you very much.

Re:Do the math

[countach74]

Perhaps you need to buy some more RAM? :)

Re:Do the math

[Anonymous Coward]

How long a while back? A while back CPUs were a magnitude slower than today, while the spinning platter HDDs were only about half the speed of current HDDs (not SSDs).

Probably one of the reasons why Intel is in the SSD market - if drives don't speed up significantly, computers will stop getting much faster for many computing tasks even if Intel's CPUs get 100x faster.

Re:Do the math

[TheLink]

And desktops started having many more cores than disks.

Re:Do the math

[Anonymous Coward]

Depends on the volume of stuff you're linking to. If you've got the RAM to read cache all the libraries you're linking against, then it's RAM FTW, but otherwise linking generates a shit-ton of random reads that an SSD will almost certainly beat disk at by a large margin.

Re:Do the math

[TheLink]

Don't put so much weight on warranty lengths.

You might want to factor in that OCZ's stuff are often even crappier than HDDs:
http://www.behardware.com/articles/881-7/components-returns-rates-7.html
http://www.hardware.fr/articles/893-7/ssd.html

Either OCZ sell defective hardware or OCZ users are many times more likely to return stuff for no good reason. I'm more inclined to believe the former.

Re:Do the math

[Shimbo]

And the thing to remember about all storage is that it will fail. If you have a single disk in a machine, and that machine is not backed up properly, then you will lose that data in the next 5 years.

The trouble is, you're probably wrong. Sure, storage will fail eventually but it's generally good enough to last the working lifetime of the machine. Plenty of home users still run XP because their machines are older than that.

Some days, I do get bit nostalgic for the days when hard drives were like that and backup/replace failed drive/restore were routine operations. Nowdays, it's a little bit more like DR: 90% chance you won't need it but then you *really* need it.

Re:Do the math

[thegarbz]

The SSD that most consumers are using are neither high end nor have such IOPS ratings.

False on the second part. Even the cheapest entry level 250GB drives these days score >90k IOPS. Performance on the cheaper drives has more than tripled in the last 3 years.

Re:Do the math

[TheLink]

Which SSDs are they using? How slow are the slowest Samsung, Plextor, Intel, Crucial, Corsair SSDs being sold today? Which of their SSDs can't do more than 10k IOPs? 10K is still 100 times more IOPs.

One "problem" I had recently was I couldn't use my SSD equipped laptop to test some DB optimization (e.g. what indexes to drop/create, which query would be better) - stuff that was slow in production was acceptably fast on my laptop (10-20x faster).

Re:Do the math

[mjwx]

9ms average access times on a 7200RPM spinning drive == ~100 IOPS.
High-end SSD: 100K IOPS.

The SSD that most consumers are using are neither high end nor have such IOPS ratings.

And this weeks Stating The Fucking Obvious Award goes to... Desler.

Consumer drives will not be as fast as enterprise drives be they solid state or spinning.

If I set up an array of consumer 10K RPM drives in my gaming boxen, do you think it will be as fast as the 10K RPM EMC boxen I have in the server room?

Of course not, nowhere near as fast.But call me when the home user has enough spare cash to drop on an entry level AX4.

You'll find that consumer spinning drives are not as fast as enterprise spinning drives.

Re:Do the math

[Anonymous Coward]

For comparison I tested my 4y old macbook against a new one bought one month ago. The new one used HD, the 4y one used a 3y old OCZ-VERTEX2. Both compiling a big iOS project turns out the 4y laptop was about 20% slower to finish compilation from scratch.

Re:Do the math

[Anonymous Coward]

If booting impacts your development cicle, you are doing it wrong (or write such crap code that crashes your whole machine!).

Re:Do the math

[TheRaven64]

And for most single-user uses, that's fine. I just kicked off a run of the LLVM test suite, since it was the first thing I could think of that would cause a lot of I/O. It was doing 1000 random reads and 500 random writes per second on my (two-year-old) laptop's SSD. It wasn't a great benchmark, because it was CPU limited, even with a quad-core i7, but that's largely the point of the SSD: the CPU is now the bottleneck, whereas with my old machine (spinning disk and Core 2 Duo), the same operation was I/O limited, even with a CPU less than half the speed.

Re:Do the math

[TheRaven64]

If you're compiling a single file, not doing a parallel build, and having all of the headers already in cache, then you'll probably see about that kind of speedup. If, however, since this isn't the 1960s anymore, you're working on a codebase with a few hundreds or thousands of files, doing a parallel build, then you'll likely see your builds take about half as long. Less if you have a faster processor, because they're now going to be totally CPU limited and not I/O bound.

Re:Do the math

[dbIII]

The links are not consistently followed. Symbolic links on NTFS are sadly untrustworthy apart from in a few cases (eg. file explorer) unless Win8.1 has done something about it.

Re:Do the math

[Anonymous Coward]

I have the same one.
I was actually blown away by how fast my computer does things now.

Re:Do the math

[mysidia]

No he's doing the math right -- At an annual failure rate of 1%, you need to replace 1% of your total capacity every year. With an annualized failure rate of 5%, you need to replace 5% of that capacity overall.

No you don't. The figures pertain to SSDs and HDDs under warranty. So you get to send 1% of your SSDs back to the manufacturer to be replaced for $0.

The figure also does not mean that 1% of your SSDs will fail

You might have forgotton the point about most people don't take adequate backups, and many people have personal photos with intangible value.

I read this instead.... if there were 1 milllion of a model of SSDs manufactured in the world in a particular year, and you bought 1, then you have 1 out of 1 million SSDs.

If the annual failure rate of that model is 1%; then 10000 of those 1 million SSDs will go bad in year 1.

You don't have a chance of losing 1% of your data from your one SSD. You have on average a 10000 in 1 million chance of losing all your data in one year.

Now in year 2.... since the annual failure rate is 1% You will have had a 10000 in 1 million chance of losing the SSD in year 1; and then a 9990 in 999000 chance of losing the data on the SSD in year 2..

Are you going to say you lose twice the cost after the end of year 2?

It's nonsensical. You will return the SSD under warranty, and receive your replacement What you lose is the data, not the SSD unit

Re:Do the math

[Anonymous Coward]

Potentially, yes. If you're interested I can sell you a custom-made infinity-TB tape storage solution. It's write-only, but you're never going to watch an infinite amount of porn anyway, so it should be perfect!

Re:Do the math

[Rockoon]

Realize that the study examines warranty claims, so there isnt such a thing as putting too much weight on warranties.

The thing with OCZ is that they have absolutely put out inferior hardware compared to their competitors. They also have, until recently, offered cheaper SSD's than all equal-performing competitor products.

Its no surprise that the cheapest item on the market is also significantly below the quality curve (the opposite would be a surprise.)

I would only trust the quality of a company that aggressively undercuts its competitors if there is an obvious reason why they should be able to. Companies in the SSD world that would have obvious advantages on margins would be Intel, Samsung, and SanDisk - each because a portion of the product is entirely produced in-house (the flash chips themselves.) Note that none of these companies are aggressively undercutting the market. SanDisk has put out some very nice deals from time to time, but they clearly would rather sell their flash chips to other manufacturers. since there is no way in hell that they could saturate their production lines without outside buyers.

Re:Do the math

[wisnoskij]

OK, and if you are working with hundreds of files, then you are compiling on a cooperate mainframe. I never tried to say that SSDs do not have a user case for compiling, but that use case is not consumer. I am physically unable to create a project, by myself (or even in a small group), big enough to warrant an SSD.

Re:Do the math

[mysidia]

The averaging is done because over time, it works out, just like insurance.

Insurance would not average in that manner.

My argument is average cost of capacity failing per year isn't a valid measurement, even if you could justify it. Price and failure rates are not constant; therefore picking an arbitrarily extreme capacity doesn't have much validity. Failure rates are higher with larger capacity units; such as 4TB drives or theoretical 4TB SSDs. Prices per capacity unit are larger with larger capacities, beyond a certain point which is media-specific.

Prices and failure rates vary with time. A SSD that costs $100 today probably costs about $50 a year from now.

The same HDD that costs $80 today probably costs about $80 a year from now.

Because most of the expensive HDD components are mechanical; there are manufacturing costs involved that won't go down. SSDs on the other hand, have most of their costs concentrated in IC and board designs that are constantly being improved and shrunk into smaller and smaller forms.

If those numbers were valid, SSDs would be coming with a much shorter warranty, because the cost to the manufacturer is enormous, since the manufacturer essentially has to pay for all the failures within warranty, to give you the free replacement....

The fact is... you don't buy 4TB SSDs, you just don't.

With 40 100gb SSDs; you might expect to have a drive fail every 4 years, with a variability of 0 to 4 drives failing in any given year, but it's much better than having 1 4TB SSD, and 0 expected drives failing every 4 years, with a variability of 0 or 1 drives failing in any given year.

The other thing is capacity is not unit neutral; the fewer units you distribute capacity over, if you maintain the same failure rate, the less likely a failure, but the higher the variation --- in other words, with large capacity units, your failure is unpredictable, the more units you distribute the same capacity over, without changing the annual failure rate, the average capacity lost increases, but the variability narrows, because your failure rate is predictable.

I would argue most people do not buy enough SSDs or HDDs for it to work out, especially not in those capacities.

It's not like you go to the store -- and buy a new SSD at the full price you paid at the beginning of the year for a brand new one; their price is not constant but a function of time, AND under warranty, the manufacturer pays this.

The valuation of capacity for the consumer is wrong, when the manufacturer bears this particular cost.

Again... your cost is the lost data and lost productivity.

These costs are maximized using units that are 4TB in capacity, if the smaller units have the same failure rate.

Re:Do the math

[Anonymous Coward]

To put this in reasonable perspective. On a not-quite current ivy bridge 4 core/8 thread machine, if I compile with sources stored on a 7200rpm HDD, I use fewer than 2 cores, because I'm disk bound. If I use an SSD, my CPU maxes out. That's the 5 times performance advantage right there.

Re:Do the math

[danknight48]

Curious,
How many libraries are you actually linking to your projects? Do you actually need them all?

Surely limiting the use of library links, and, optimizing your program in the process. Is more important to productivity than how fast your HDD can link unnecessary library links?

Linking a game engine as a .dll:
My current code size is 2mb in the project, and, it takes me 2 seconds to compile on a "trusty rusty".

Re:Do the math

[Anonymous Coward]

More to the point, you can buy 4 4TB HDDs for $800 and setup a RAID1 and get a lot of the same read performance as an SDD while having heavy redundancy.

No you can't. 1) RAID 1 doesn't improve write times at all (in fact, it minutely slows them as you have to wait for all drives to finish writing). 2) SSDs are 3 orders of magnitude faster than HDDs at random reading, your paltry half an order of magnitude reading doesn't come close to equalising that.

Re:Do the math

[TheRaven64]

I do, and so does everyone else on my corridor. Yes, we also have machines in racks that can build things faster, but with a quad-core i7 and an SSD in my laptop, I can build the LLVM codebase in 5-10 minutes, with incremental builds often taking under a minute, and that's a big win for my productivity over spinning rust. Most software projects that I work on have been ongoing for 10 or more years and in that time even a small team can produce something that has thousands of source files.

Re:Do the math

[wisnoskij]

OK, so it is a consumer use-case, because compiling is so fast that it is not worth using the companies machines, but it is slow enough to pay a few grand to half the compiling time? Why not just network to a central repository on a super fast, SSD carrying, mainframe.

Re:Do the math

[TheRaven64]

I have a build machine that has 24 cores, 256GB of RAM, 256GB of SSD, and 4TB of spinning rust. Building on the SSD is significantly faster than the spinning rust (factor of 2-4 depending on what you're building). Building on the RAM disk is probably 1-2% faster, but within the noise for subsequent runs. I often put the object files on the RAM disk though. It doesn't seem to impact performance, but it means that the OS never has to bother writing them out to disk, which should help the SSD last longer when I'm generating 2-30GB of object files per build.

Re:Do the math

[TheRaven64]

Uh, what? Compiling is so fast on an existing, consumer-grade laptop with an SSD that for incremental builds it's a productivity boost, and it works when I'm at home, at work, on the train, or whatever. The difference in price between the SSD and a mechanical disk was about £300, two years ago (for a 256GB disk). It's now even lower. When I'm doing a lot of builds in different configurations I'll use a centralised server. When I'm actually hacking, I want my working copy to be local. The entire point is to speed up the edit-compile-test cycle, and sticking a network in the middle does not do that.

Re:Do the math

[h4rr4r]

You will however find that consumer SSDs beat the ever living shit out of any Enterprise SSD. Actually I have tested consumer vs enterprise hard drives and they are pretty close. That EMC box is doing a lot of magic to hide how bad hard drives are.

Re:Do the math

[h4rr4r]

A few grand? You can get 128G SSDs for $100.
If you have a laptop simply their G load tolerance makes SSDs a no brainer.

Re:Do the math

[h4rr4r]

How big are yours? Ram is cheap. You can get normal middle of the road laptops with 16GB of it.

Re:Do the math

[Anonymous Coward]

There's enough software out there that simply doesn't see/follow NTFS symlinks that it's not a reliable option. (How it doesn't see/follow the link is beyond me, since that *ought* to be handled fully at the filesystem level, but it happens.)

Re:Do the math

[h4rr4r]

So one guy who failed to to proper benchmarking before is your evidence?

If you don't know what your bottleneck is going out to buy hardware is a stupid move.

Re:Do the math

[Anonymous Coward]

Realize that the study examines warranty claims, so there isnt such a thing as putting too much weight on warranties.

Please read more carefully. What was said was "Don't put so much weight on warranty _lengths_".

And that was clearly in response to your post. e.g.

You might want to factor in that SSD's often have _longer_ warranties than HDD's these days.

(emphasis mine in both cases).
Longer warranties are not much comfort to the users if the product has a 20% or higher failure rate (e.g. OCZ). Most people buy SSDs to use them, not to keep returning them for years.

As for OCZ the term defective seems more appropriate than "inferior" given the failure rates.

Re:Do the math

[Gibgezr]

Actually, it's the same in Windows. I dunno what k8to is doing, but for most people it's the compile that takes a lot longer; linking is relatively quick.

If you just stop and think about was work the compiler has to do, versus what work the linker has to do, it makes sense.

Re:Do the math

[kumanopuusan]

I'm generating 2-30GB of object files per build

I'm not sure if I should be impressed or disgusted. It's actually a little of both.

Re:Do the math

[mlts]

Another idea is to mount a partition using a junction point (Windows's analog to UNIX mount points.) That way, it might be happy storing its stuff under c:\Program Files\blarf, while directory blarf is actually a different partition.

I do this for Cygwin which makes so many small files that a chkdsk takes forever, so the Cygwin stuff goes on a different partition.

Re:Do the math

[TheRaven64]

A debug build of LLVM generates over 2GB for the debug info. Much of this is removed by the linker. For FreeBSD, it's the userland for 6 architectures when I'm doing a test build and a few dozen kernel configs. It adds up quickly.

Re:Do the math

[Rhipf]

So you would rather spend $250 for a 150GB SSD drive instead of $60 for the smallest magnetic drive you could get (lets assume 320GB) just so you can fill the drive and not feel like you are wasting storage space?

Sorry, I don't follow the logic.

Re:Do the math

[MikeBabcock]

This is not flamebait, its correct:

5% of a 4TB HDD that sells for USD$200 is roughly 200 GB = $10.

1.5% of a 4TB HDD that sells for USD$29,000 is roughly 60 GB = $425.

If you don't understand why, read up on annualized failure rates. Someone get out your informative mod and read the parent.

Re:Do the math

[MikeBabcock]

Your entire post is nonsensical. Nowhere in my post or the gp to my post did we assume anyone loses data or doesn't do backups.

I do drive allocations for enterprise customers, we sell dozens of drives, not one or two. We make people do daily off-site backups. We enforce RAID-1 minimally for all storage. This all comes from understanding annualized failure rates on drives, which is what I posted.

Once you understand how drives fail, and what it will cost you to replace them, you can plan ahead for those costs.

Understanding *that* they fail at all means using RAID for high availability systems and backups for all systems.

Re:Do the math

[MikeBabcock]

Yawn, your posts are boring me because to some of us the math is too obvious to be bothered explaining to others.

It is in fact *exactly* how insurance companies work. They know they need to plan ahead for N payouts of X dollars, that a specific insurer has an M percentage chance of making a claim of Y dollars, and charges them Z per year to cover that annualized cost of those possible payouts (plus profit in their case).

If you're not doing the same math for hardware replacement, especially in business, you have no business making computer decisions.

Re:Do the math

[Anonymous Coward]

And none of this build/compiling is being done on a dedicated build server? How cheap is your fscking boss to not have a dedicated build system? Hell a dedicated build system is pretty fscking cheap now - in fact it's cheaper then what even an noob dev makes.

Re:Do the math

[Anonymous Coward]

Why are you booting so often? Sleep mode works quite well in Win7 and even hybernates properly for when the power craps. It can't be for any kind of power savings because the standy 5v power demand of any system is 1w even when the fucking power is off. The only way to cut that parasitic load is to use a power strip and physically cut the power to the system, monitor, speakers and anything else you have connected to it.

When I used Win7, I never rebooted my system except for power outages (battery backup solved most of them) and Patch Tuesday. Otherwise, I had the damn thing dropping into sleep mode in 10 minutes. Hybrid sleep didn't kick in until system had been idle for 30 minutes (meaning I wasn't home).

Fast Turtle

Re:Do the math

[Mad Merlin]

Most workloads are in fact dominated by small, mostly random, reads and writes, which is why SSDs are just that much faster in the majority of cases.

If you're talking about mainly sequential reads, then the situation for the four RAID1 HDDs is even grimmer. RAID1 provides virtually no speedup for single reader sequential reads, as to do so would require tons of seeks from the drives (which as we know, HDDs fail at), or an extremely large file and very large stripe size (and also a matching amount of memory for intermediate buffers). Most RAID1 implementations don't even bother trying.

Having said that, HDDs are substantially better at sequential reads and writes than random ones, and if your workload really, truly is dominated by sequential operations (and it probably isn't), you can generally match the performance of a single SSD with a RAID10 of roughly a dozen HDDs (or a RAID0 of half a dozen, but say goodbye to reliability). This ignores the fact that a dozen of even the cheapest HDDs is substantially more expensive than an SSD, due to actual unit cost, the extra power draw, the extra physical space required for them, the extra HBA(s) to plug the drives into, the extra manpower to install/manage them and the extra manpower to deal with them when they die.

There are still reasons to use HDDs, but performance is absolutely not one of them. It's not even close. Take it from someone who manages several hundred HDDs + SSDs.

Re:Do the math

[fnj]

Are they really? Not in my experience. Symbolic links are most decidedly not "shortcuts".

Re:Do the math

[fnj]

Why did you not use a symbolic link to another drive and copying the entire /User tree using xcopy or Explorer?

Re:Do the math

[kumanopuusan]

Ah, that's pretty reasonable compared to whatever I was imagining.

Re:Do the math

[KiloByte]

If you're really a developer, you'll see your compile times cut by a factor of 5-10

Please tell me, in what circumstances sources you're building are not already in memory?

And even if they were not, just increase the -j level by one so no CPU time is lost waiting for I/O. So both builds will be CPU-bound.

Sorry, but if your compile times change noticeably, much less by a factor of 5-10 as you say, you're doing it wrong.

Re:Do the math

[KiloByte]

here's another: when I switched to an SSD at work my compile times were cut by more than half

Here's a -j switch for you. Please learn how to use it. If any part of the compile is waiting for the disk, you'd better fix your tools.

Re:Do the math

[Anonymous Coward]

Which means nothing because the large size means those few write cycles are still WAY more than most people will ever write to it.

Re:Do the math

[WuphonsReach]

that's largely the point of the SSD: the CPU is now the bottleneck

Yep, it moves that bottleneck around. I have an older Core Duo chip in my 2007 era laptop and I can tell when I'm CPU-pegged now. It's the only thing I don't like about my Thinkpad right now (8GB RAM, Win7, 300GB SSD). But everything else works perfectly, so I put up with the slower CPU and don't plan on upgrading until 2015 or 2016.

(At which point my laptop will be 8-9 years old.)

It helps that I have a 4-core Linux server and an 8-core Win7 desktop that I can use to offload CPU-heavy stuff to (such as video encoding / transcoding).

Re:Do the math

[toddestan]

That may be the case if you're compiling something on an old PC that has 128MB of ram or something. But nowadays when I have 10GB+ of ram sitting around that can act as a disk cache, it's a non-issue.

Re:Do the math

[brit74]

If you don't know what your bottleneck is going out to buy hardware is a stupid move.

He was upgrading to SSDs, so he did a test before and after the upgrade. He didn't say that he was upgrading specifically to speed up compiles. He also said that everything else worked much faster on his computer, so he was happy with the purchase. (So, your criticism about how "stupid" he is for upgrading kind of misses it's mark.)

Re:Do the math

[mysidia]

Understanding *that* they fail at all means using RAID for high availability systems and backups for all systems.

RAID is not adequate for SSDs, because they have failure modes that are related to characteristics of drive longevity and created by the pattern of wear (interaction with wear-levelling on specific models) -- number of writes over time.

That is; correlated failures between RAID mirrored pairs are more likely than not with SSDs used over long periods of time.

Therefore: with SSDs, duplicate data makes sense, and RAID does not.

Re:Do the math

[Dahamma]

Of course we have build servers - dozens of them running incremental builds in every checkin, as our platform is running on hundreds of different devices. But this thread is about SSDs in workstations for development - for developers who build and test their changes before checking in code (many of which may not have dedicated servers anyway).

Re:Do the math

[MikeBabcock]

Nobody said the RAID set would be two SSDs.

cf. http://tansi.info/hybrid/

Re:Do the math

[10101001 10101001]

Most workloads are in fact dominated by small, mostly random, reads and writes, which is why SSDs are just that much faster in the majority of cases.

Yes and no. SSDs offer much lower latency and can handle near instantly the common small, mostly random, reads (and writes, but that's heavily covered with RAM caching). But the common workflow rate is nothing closer to on the order of 90,000 IOPS. Instead, the issue is more that spikes in most workflows can reach the low thousands range and HDDs will strain for seconds to fulfill requests*. Hence, the actual performance difference is on the scope of SSDs being 5-10x faster. It also heavily implies that with a proper RAID1 implementation ferreting out requests that a lot, though not all, of the read requests could scale near linearly across all the drives. The main issue, as you note, is that plenty of RAID1 implementations likely don't do such things well if at all (although I wonder if what you say holds true for Linux's dm-mirror (which I know isn't strictly RAID1, and you can readily fault me that I wasn't necessarily strictly speaking of the official spec RAID1 but primarily of a fault-tolerant mirroring system)).

There are still reasons to use HDDs, but performance is absolutely not one of them. It's not even close. Take it from someone who manages several hundred HDDs + SSDs.

Except my point wasn't that HDDs are better performance wise, even in a RAID configuration. It was that a RAID of HDDs could obtain a better fault tolerance rate and still see a substantial performance spec of SSDs (really, you should look at the benchmarks I linked to before or try to find other workflow examples to see just how big of a difference in scale having 1000x the possible IOPS really has on most use cases) with the unstated point of relatively cost being a lot less--and 4TBs were admittedly a bad example because 1TB (on the order of $60/drive) are more than sufficient for a comparison given the SSD price range for such things. I don't think your notes about power usage hold up--4 6W drives** work out to be ~1kW*h/1.736 day sustained or ~$31.53/year at $0.15/kW*h--figures I think are just overboard to realistic. Maintenance and enclosure costs are the real killers, I think, although the former has to be weighed against the downtime costs on an SDD system vs a RAID system and the latter can be possibly avoided (presuming maintenance is done after hours and computers were bought with sufficient internal space for the necessary drives--admittedly a possible blocking issue).

In any case, yes, SSD is the clear winner as far as raw performance goes. But the whole article was about reliability. And you and others keep bringing up IOPS as if they remotely linearly scale to real world performance (they rarely do). Meanwhile, RAID HDDs offer a pretty clear middle ground for a lot of people who care about reliability first and can still see performance gains in many cases. But, I'd readily defer to your experience upon the subject. I'd just prefer some actual numbers to back it up rather and some clear observations to sustain that position than using vague words like small and big or random and sequential or focusing on IOPS when throughput is a major factor too in judging things too :)

*I base this all on personal observation of actual throughput on systems which varies greatly but can readily sustain in the 7MiB/sec to 20MiB/sec range which based upon 120IOPS implies minimally common sequential reads/sec of at least ~60KiB to ~170KiB blocks (presuming saturated IOPS) and likely much higher actual numbers since I doubt IOPS requests are actually commonly saturated. Of course once you do get to saturated IOPS on a HDD you are looking at a potentially dismal (presuming 4KiB/block) 480KiB/sec while SDD is likely only limited by its ability to saturate SATA or whatever the connection is. Yet benchmarks rarely bare out figures like that (that I've seen, anyways)

Re:Do the math

[mysidia]

Yawn, your posts are boring me because to some of us the math is too obvious to be bothered explaining to others.

You mean that it is too obviously wrong.

It is in fact *exactly* how insurance companies work. They know they need to plan ahead for N payouts of X dollars, that a specific insurer has an M percentage chance of making a claim of Y dollars, and charges them Z per year to cover that annualized cost of those possible payouts

The point is, "that a specific insurer has an M percentage chance of making" IS NOT THE SAME QUANTITY as the percentage of the insured that will make a certain claim; the two numbers are not interchangeable, that is to say: they cannot be substituted for one another.

Probability of a single member of group G failing not equal to The proportion of members of G expected to fail.

And when we have small number of SSDs..... well.... nobody buys "4TB" SSDs; and I never heard of anyone buying "4TB" HDDs, if they even make them. Both of those things are obviously so contrived.

Re:Do the math

[Mad Merlin]

I run about 90% of the systems I manage in RAID 10 (there are a few oddballs in there, some only support 2 drives, those are RAID 1, and there's a few where I don't care about performance, but do care about drive space, those run RAID 5/6). The real world performance difference of RAID 10 over a single drive is very large. Assuming a four drive RAID 10 array, expect between 2x and 4x improvement in both random and sequential read/write performance.

WIth that in mind, at $dayjob, we run a lot of VMs. Before SSDs were affordable, we could usually fit between 6 and 8 VMs on a single host (with 4x or 6x 7200 rpm drives in RAID 10) before they became unusably slow, with tons of time spent in disk wait. CPU time and memory usage were rarely limiting factors. As soon as we started deploying SSDs, the only problem was running out of space. Right now we have over 50 VMs running on a single 8x SSD RAID 10 array, and it's blindingly fast.

There's a similar story with databases. Back before SSDs were affordable, we bought a machine with enough RAM to keep the entire database cached in memory, as it was just too slow to run off of 15k RPM SAS drives. On a fresh boot, we'd still need to precache the database into memory, and with said HDDs, that's a job that took something like 10 minutes and was almost entirely disk bound. We recently upgraded that machine to SSDs, and the same precache task now takes under 30 seconds.

As for home users, well that's a different story. Personally I think it's downright irresponsible to run any system with a single drive (HDD or SSD), but the overwhelming majority of existing machines with a single drive suggest that my opinions on this matter are not widely held.

I guess my issue with your proposal is that I just can't see very many cases where it's practical. The low end of the market is dominated by Laptops/Desktops/Tablets/whatever that cost under $500 and all have only a single drive, as an extra $100 for another drive is going to be a dealbreaker most of the time (if another drive would even physically fit). The high end of the market where performance is critical, is completely dominated by SSDs. You can read countless stories of big companies replacing full racks (42U) of HDDs with 1U or 2U of SSDs. I guess somewhere in the middle there is a small set of people who:

• store a lot of non-media* files (over 500G or 1T)
• are not overly concerned with performance
• have the technical know-how to set up and maintain a RAID array
• are significantly more concerned with reliability than most
• are still relatively cost-sensitive

Those people would probably be better served by a 4x HDD RAID 10 array than a 2x SSD RAID 1 array.

* If you're storing media files on SSDs, you either have too much money to burn, or zero sense. They're huge and 99% of the time are read/written sequentially.

Re:Do the math

[greg1104]

100 IOPS is an accurate number for 7200RPM drive. The 100K IOPS numbers some SSD vendors trot out are best case fantasy.

Last week I watched a 400GB Intel DC S3700 hit its limits for a long period. Intel is conservative on its numbers for this drive compared to a lot of the numbers other companies advertise. They only claim 36,000 IOPS. But here's what a real world worst case for the drive looks like, from a busy PostgreSQL database server:

Device: rrqm/s wrqm/s r/s w/s rMB/s wMB/s avgrq-sz avgqu-sz await svctm %util
sdc 0.00 22.80 888.00 483.00 6.82 3.96 16.10 162.34 118.79 0.73 100.00
sdc 0.00 2.60 908.00 422.00 6.27 3.32 14.77 170.80 125.76 0.75 100.00
sdc 7.00 54.20 1102.00 1900.40 11.64 15.26 18.35 166.82 56.68 0.33 100.00

I tell people they shouldn't expect more than 1500 IOPS on real mixed workloads with lots of concurrency from any single SSD. That's great, and it's completely reasonable to say a SSD is always going to be 10X faster than a mechanical drive. But the only devices consistently delivering even a real 100X speedup are the super expensive models from companies like FusionIO, Virident, etc, with matching price tags. And nobody is really doing 100K IOPS in anything but carefully controlled lab benchmarks.

Re:Do the math

[10101001 10101001]

I guess my issue with your proposal is that I just can't see very many cases where it's practical.

Granted, most desktop users buy a computer as a unit and don't make modifications. Those that do rarely know ahead of time what they want and even if options were available by OEMs, a lot would make poor choices. So, it seems an almost inevitable that only the technically minded who are willing to make modifications are the majority in the middle, without the cash to spend the significantly more on a pure RAID SDD setup. Of course, the same people are liable to just use one SDD as a boot/root device and forgo the idea of RAID. :) And the practical SDD/HDD hybrids which I can only presume have even worse reliability than standard HDDs or SDDs (I don't presume that hybrids do good fall back onto just SDD or just HDD if one part fails) will likely be the major penetration in the desktop arena.

Still, I'd prefer a 4x HDD RAID for the performance/capacity/reliability (the latter is more important to me actually as I've seen plenty of computers die). But, then even I'm in the boat of having a system not designed to physically contain that many HDDs and my system doesn't support eSATA or USB 3.0 for an enclosure. :( So, it's all still too much of a theoretical for me. It's nice to think about The cost of multiple SSDs isn't, though.

Re:Do the math

[Man Eating Duck]

Yes, sure you are right: didn't use one because I never felt the need for an SSD while dev-ing. Certainly, the fact I'm not required to use Visual Studio or the like helps.

In addition to better performance on some IO-heavy tasks, what makes even a cheap SSD stand out is close to zero access time. Everything is more responsive, large programs launch in literally 1/10 of the time. The performance difference is slightly less noticeable in Linux, but it's still a huge boost. The system is simply more pleasant to use. As drkstr1 says, it's obvious that you've never tried one :)

Every carpenter knows that it's worth it to invest in quality tools, I'm surprised that a developer who presumably does serious dev work would skimp on a crucial part like the system drive. You could probably get by with only 2GB of RAM as well, but why on Earth would you want to do that?

A musician I know who also does a significant amount of studio work asked me if I could help him setting up a Linux computer for mastering. He preferred using a Mac, but would like to save some money going the Linux route when getting a new computer. I advised him to fork out for the Mac instead, it would not be smart for him to reduce the productivity and enjoyment of a significant portion of his livelihood.

Re:Do the math

[Man Eating Duck]

I've been running multiple partitions/drives on Windows for years and haven't encountered a sizeable app that I'd want to put on a different drive that wouldn't let me.

Same here. I've also found that for gaming Steam Mover makes it very easy to shuffle my current games onto the small SSD. It works for any directory, not just Steam games. Windows doesn't know the difference, but some games benefit greatly from living on the SSD. It would likely fool any stupid software that insists on living on C: as well :)

Re:Do the math

[Rockoon]

sigh...

The study is based on warranty claims, which do not happen outside their length. Its important to note the differences in warranty lengths precisely because of that fact.

You seem to think that warranty lengths are not important to the study and the 5% and 1.5% figures... what a fucking retard you must be. The devices with the lower claims rate have longer warranties... you seem to want to insinuate with hand waving that somehow the opposite is true. Dipshit.

Re:Do the math

[TheLink]

I'm actually more of a jerk/asshole than a dipshit. But you're intentionally continuing/pretending to be stupid.

The devices with the lower claims rate have longer warranties... you seem to want to insinuate with hand waving that somehow the opposite is true.

Where did I insinuate the opposite? I merely said you shouldn't put so much weight on them, while you said "the study examines warranty claims, so there isnt such a thing as putting too much weight on warranties."

You yourself mentioned OCZ's SSDs (including the Octane) have 3 year warranties.

Corsair SSD warranties are 3 years - return rates about 1%.
Most of Crucial SSD models have warranties of 3 years. Return rate a little over 1%.
Samsung SSD warranties range from 3 years to 5 years, and the 3 year ones (e.g. 830 ) definitely don't have a return rate anywhere close to OCZ's crap or it would certainly be noticed by now.
HDDs with 3 year warranties still have lower return rates than OCZ's 3 year warranty SSDs some of which have return rates of up to 40%. NONE of the HDDs in the report I mentioned have return rates over 10%.
In contrast OCZ's crap have return rates of 9%, 30%, or even 40% while still having 3 year warranties.

Hence you shouldn't put so much weight on warranty lengths. QED.

You had a chance to stop being stupid in your reply but you clearly chose deliberately to remain stupid.

For the record, you did the insulting first, so you're as rude and offensive as I am if not more so, in addition to being _voluntarily_ more stupid/incompetent, hence you've proven yourself to be the real dipshit in this thread[1].

[1] You might not be a dipshit elsewhere of course.

Re:Do the math

[Decker-Mage]

I've argued with my fellow economists about "diminishing returns to scale" and so far the jury is still out, but seriously leaning my way as some of my colleagues have conceded . Significantly, so long as Moore's Law still applies, we have increasing returns to scale when it comes to capital investments, especially when factoring in intellectual capital. Each generation of tools gives us significantly better capabilities for any particular level of capital investment in a (virtuous?) positive feedback loop.

I've been using SSD's since the first generation and I like them. Each generation is cheaper, more capable, and more reliable, not that I've ever had one fail (knock on wood) despite using refurbished drives for my throw-away scratch. And given that they are used in a scratch configuration (VM's from golden images, large/big data, ...), they should have a significant failure rate here. OTOH, hard drive failures are fairly common which is why I get the extended warranties for them, not the SSD's. Anecdotal evidence, but evidence nonetheless.

But the disc can store much more

[Marrow]

So you need to multiply the failure rate of the SSD by as many SSDs as it would take to equal the storage of the disc. Do you want the storage rate per arbitrary device size, or rate of failure per data stored?

Re:But the disc can store much more

[Rockoon]

Thats a silly thing to do. Lets examine this, shall we?

A 5% chance to lose 2TB vs a 1.5% chance to lose 250GB.

You argue that since it requires 8 of these 250GB SSD's to equal the capacity of the 2TB HDD that we should multiply 1.5% by 8, so a 12% chance... a 12% chance of what, tho? In actuality, there isnt a 12% chance of anything...

The chance of losing at least 1 of those 8 SSD's (that is specifically 1 or more) over the period is (1 - (1 - 0.015)) = 0.114, but the chance of losing all of those 8 drives over the period is 0.015^8 = 0.0000000000000025628906. In other words, losing all 2TB in the SSD scenario is effectively never going to happen while it remains 5% for the HDD scenario.

The actual breakdown of all possibilities of drive failings (0 drives, 1 drive, 2 drives, etc..) rounded to thousands of a percent is:

0 drives: 88.611%
1 drives: 10.795%
2 drives: 0.575%
3 drives: 0.000%
4 drives: 0.000%
5 drives: 0.000%
6 drives: 0.000%
7 drives: 0.000%
8 drives: 0.000%

So we see that you would be twice as likely to lose some data than in the HDD scenario, but invariably it will only be 250GB of data instead of 2TB of data (only 1 in 173 of these 8 drive experiments will witness more than 1 drive fail, and the majority of those will be exactly 2 drives failed)

So no, you do not need to multiply the failure rate of the SSD's by the number of SSD's that you would need to equal the HDD. What you need to do is define the problem better because as it stands SSD's look a hell of a lot better when you suppose that you need a pile of them.

Re:But the disc can store much more

[Rockoon]

Woops. Missed an exponent when typing that out. The first equation is supposed to be (1 - (1 - 0.015)^8) = 0.114.

Re:But the disc can store much more

[hedwards]

If you've properly backed up the data that's a moot point. And I can't recall ever having had a HDD die so suddenly that I couldn't make one last image before it bit the dust. Whereas I have regularly had SSD of various types die over the years in such a way that none of the data could be recovered.

Also keep in mind that those 5% of HDD that failed typically failed in a way that you could still access the last data that was on there. Even if you had to have a specialist remove the platter and read it. A SSD that goes bad means that damn near everything is lost due to the wear leveling that's built in.

I'm sure it's possible to have a severe head crash or somehow shatter the platter resulting in a disk that goes bad instantly, but I've never seen it happen and I don't think I've even met somebody that's had it happen to them or known anybody that's had that happen.

Re:But the disc can store much more

[Anonymous Coward]

You got a lot of book learnin' but short on common sense.

The failure rates may not be completely independent if you have them in the same computer, in the same conditions, and the same or similar usage patterns. And will *definitely* not be independent if they come from the same NAND manufacturer (of which there are not very many in the world), and SSD controller and FTL developer.

Re:But the disc can store much more

[Anonymous Coward]

So what you've really told us here is that: you are twice as likely to lose data with a SSD as with a HDD. Sure, you'll only be losing an eight of the total amount of data, but you're chance of failure is now much, much higher. Even double. Too bad we won't get to pick which 250GB we lose. Could be empty space, could be something important.

Re:But the disc can store much more

[Anonymous Coward]

But it's enough that you loose any single one of the SSD's to loose the data.... And in my case i prefer to have a single large volume ( ie SSD would have to be in a raid ).
The problem with having consumer SSD's in raid's are quite a few.. I have experienced disks returning corrupt data without reporting anything bad and all sorts of things..

But to sum it up.. A single HDD of 2Tb is less likley to fail than 8 raided 250Gb SSD's and 8 raided 250Gb SSD's might cause silent data-corruption in some cases.

But the second thing that nobody seem to notice is that SSD's do fail quite fast depending on the workload...
1. Build-server - large repository of source + many builds per day = lots of small writes.... Since the erase-size of a SSD is usually 256Kb quite a few small writes will cause erases to be executed and there by wearing the flash down. SLC helps but does fail..
2. Developer - large repo + many builds per day.. Not as much as a build-server but still.. They also do small changes to small files causing small things causing erases of larger areas. we have a ~1.5 year estimate on a 500Gb that is not allowed to be filled to more than 50%. (MLC drive)

For both of the above a HDD + lots of ram is much better.. (extra ram is only for caching writes and disk-cache)

Buying RAM for the extra money a SSD costs usually gives more performance for most people. (ignoring start-times of applications etc)

Re:But the disc can store much more

[Anonymous Coward]

... but the chance of losing all of those 8 drives over the period is [snip]

Which is completely irrelevant.

Loosing one SDD drive outof a series has pretty much the same inpact as loosing all of them. Your data integrity is gone, just as with loosing some random sectors on a HDD.

If you really want to compare it, than do it right please.

Re:But the disc can store much more

[Anonymous Coward]

How many FS/partition tables/bootloaders on those SSDs ?
How many random 250GB-big holes in your FS before it becomes a 2TB data loss ?

Re:But the disc can store much more

[rnturn]

Wow... I'm reading these comments (not just yours) and I'm a little flabbergasted as to how blase people are about losing data. People say things like "Well... I only lost 250GB" implying that it's not big deal and I'm flashing back to a time only a few years ago when that would have been a significant chunk of a corporate data warehouse I managed. I can't imagine the look on my boss's face if I had to tell her that "Hey, it was only 250GB". Sure we could restore that from backups but the downtime would have been inexcusable if this much data had just been lost without warning.

Re:But the disc can store much more

[KiloByte]

You're describing a JBOD setup then analyze its failure rate as if it was an 8-way RAID1. GP's point stands: the true failure chance is 11.3%.

Without any kind of replication, any disk failing means your system fails immediately. If, on the other hand, you had those SSDs in a RAID, a fair comparison would require a RAID on the HDD side as well.

Re:But the disc can store much more

[Anonymous Coward]

Crucially though, note the use of the key phrase "failure rates under warrenty". Looks like nearly all SSDs are going for 3 year warranties. Hard drives appear rather more variable at this point, but 3 year is the most common though quite a few still offer 5 year warranties and a few offer 2 year. Given the extremely sharp failure curve on Flash, I'd guess they'd have better than 50% failure rate at 5 years, while hard drives might hit 10% in that timespan. Seems like the hard drive manufacturers need to tighten things down so they can offer 5 year warranties on everything since that is when they win versus SSDs.

Re:But the disc can store much more

[Anonymous Coward]

You assume total independence of the failure chances, while in fact there is very strong correlation due to common mode failures.
* Similar load causes the flash to wear out at approximately the same time
* Bugs in the firmware
* Some models or model versions or batches are highly prone to failure.
* Power supply outputs out-of-spec or unstable voltage
* Controller dies
* Virus exploits firmware vulnerability
* Bugs in OS that introduce unexpected load
etc

Re:But the disc can store much more

[Anonymous Coward]

why would you do that, this is a failure per device not per bit argument. I am sure that you would also have more failures per floppies, and less per dvd.

Great...

[Desler]

Yet, according to IHS, client SSD annual failure rates under warranty tend to be around 1.5%, while HDDs are near 5%. So an SSDs not only outperforms, but on average outlast spinning disk.

What about annual failure rates outside of warranty?

SSD failure rates

[rjr3]

If you have a new Apple notebook it does not matter what the rate is - you can not replace them.
Lose the SSD and you have lost the Retina.

Re:SSD failure rates

[haruchai]

Yet another reason to never buy Mac.

Re:SSD failure rates

[BronsCon]

Off-topic but... I bought my wife a new MacBook Pro a couple months ago because I couldn't afford one in 2011 when they last made the 17" models and they haven't made a laptop worth buying since (the MacBook Air is a netbook... and netbooks were never worth buying). I took her late-2011 17" and love it; she always said it was too bulky, so it was a win for both of us.

To bring it back on topic: I've killed quite a few hard drives; I seem to lose one every year, on average, and have for as long as I've owned a computer with a hard drive (that is to be read as I've lost as many 5yr old drives). With a 1:1 ratio of SSDs to spinning disks in my home for the last 2 years, I'm still losing HDDs at the same rate, but the SSDs seem to be holding up. I don't mind the diminished capacity; that's offset by the extremely improved performance and replacing the DVD drive with a HDD caddy for bulk file storage; combined with network storage (and VPN for remote access) and automated backups, I still have all of my data.

Re:SSD failure rates

[gander666]

Bullcrap. They can be replaced. Look up http://macsales.com/ they sell several sizes for the airs and the pro retinas.

Re:SSD failure rates

[Anonymous Coward]

People with Macs are professionals who have more important shit to do than ementary pc technician bullshit, just take it to the Applestore and you get a new one and get back to work.

Re:SSD failure rates

[Anonymous Coward]

How dare you bring facts into this conversation!

Re:SSD failure rates

[drhank1980]

Not completely true. You can get one from OWC . You will pay more for it and it is a pain in the butt to do the work but you can replace them. I do remember when working on mac hardware was easy and quick but those days are long gone.

Re:SSD failure rates

[Anonymous Coward]

The SSD may be replaceable, but it uses an Apple proprietary interface which limits options. Furthermore, how about upgrading or replacing failed RAM? In that case, you are left with a $2200+ brick. (not even usable as a doorstop, thanks to Steve's obsession with thin at the expense of functionality...)

Re:SSD failure rates

[Anonymous Coward]

Yeah, after you buy that DIY kit. A laptop hard drive replacement shouldn't require anything more than a normal screwdriver and the removal of a single panel. With the Macbooks, you have to do a complete teardown with specialized tools.

Re:SSD failure rates

[dfghjk]

"The SSD may be replaceable, but it uses an Apple proprietary interface which limits options."

How is that remotely similar to "Lose the SSD and you have lost the Retina."?

"Furthermore, how about upgrading or replacing failed RAM? In that case, you are left with a $2200+ brick."

Just like hundreds of other components in PCs throughout time. Pure FUD, nothing more.

Re:SSD failure rates

[dfghjk]

Boo hoo.

Re:SSD failure rates

[BitZtream]

Really? Thats funny because I did.

Perhaps a cluepon would be something you would look into.

Apple will replace them under warranty of course, or you can pay out of warranty.

You can also buy them from OWC if you want an alternative. http://eshop.macsales.com/shop/SSD/OWC

Re:SSD failure rates

[hobarrera]

How is a MacBook Air a netbook? An i5, 8gigs rams, SSD, I can plug it into my monitor when I get home. It's also as powerfull as medium-grade desktop. What's is missing?
I hate to bring it to you, but an MBA is exactly like any other ultrabook out there.

Re:SSD failure rates

[BitZtream]

Still is if you don't want the thin models. They still offer traditional units with user serviceable ram and drives, just like every other laptop thats thick.

Of course, show me the ultra book thats as thin as a MBP with replaceable RAM ... hint: You won't, they do the same thing.

Re:SSD failure rates

[Maxo-Texas]

You may want to look into a power conditioner.

My laptop drive is 7 years old (runs XP).

My desktop drive is close to 5 years old.

I use them a lot.
They were on for about 3 years solid tho I've been putting them to sleep the last year.

Your failure rate seems suspiciously high.

I also have several USB drives of similar ages.

The only drives I've ever lost were 3 flash drives. Two of them mini drives which got very hot during use. And an old 88 mb drive back in the 90's. (cost me $88!)

I still back up frequently.

Re:SSD failure rates

[Anonymous Coward]

You're full of crap -
I've personally replaced a number of drives, and upgraded a bunch.
The ssd's are readily available. Upgrading to the 768 SSD is quite popular for those on smaller sizes, especially when its not at Apple original pricing.

Gets a bit more convoluted for the Air's, as there are several designs now, but you can still readily get replacements.

Re:SSD failure rates

[mlts]

I second the recommendations for a power conditioner, or perhaps an online (not standby) UPS [1]. You can make your own online UPS with a Victron or Magnum Energy "hybrid" inverter/charger and a couple 12 volt AGM [2] batteries. Clean power can make a big difference in drive and component life.

[1]: An online UPS always uses the batteries, a standby only switches to the batteries when mains power is out. I'm sure one can guess which will help with voltage sags and other dirty power issues.

[2]: AGM batteries are better since they don't outgas, so can be used indoors.

Re:SSD failure rates

[WuphonsReach]

You may want to look into a power conditioner.

Or the quality of the PSU inside the case. Bad PSUs are probably the #2 enemy of spinning hard drives or any electronics (heat is the #1 enemy).

Re:SSD failure rates

[Anonymous Coward]

People with Macs are professionals

The opposite is true. The people who have always bought Macintosh are clueless amateurs who can't figure anything out.

Re:SSD failure rates

[Maxo-Texas]

Yea, I'm not crazy about it but I do clean the inside of my case of dust roughly once a year.

Re:SSD failure rates

[BronsCon]

You're going to assume that I have poor power quality rather than a large number of drives. How cute.

SSDs are just as prone to controller (or other chip) failures as spinning disks, and those are the failures most likely to be caused by poor quality power. If that were the cause, wouldn't I be seeing SSDs fail at the same rate? I'm not quite sure how you arrived at your conclusion; however, I can assure you, as I glance over at my power distribution rack, that it is incorrect.

Re:SSD failure rates

[BronsCon]

Did you notice that netbooks disappeared when ultrabooks hit the market? Having owned an Acer Aspire One (netbook) and currently owning a Sony Vaio Duo 11, I can tell you that, despite having 3x the RAM, an 4x the SSD storage (the One had an SSD as well), and 4x the CPU cores running at 2x the clockspeed, the only functional difference is that the Vaio collapses into tablet mode.

Ultrabooks are netbooks, my friend. Very. Expensive. Netbooks.

Re:SSD failure rates

[BronsCon]

Oh... I forgot... The One could connect to a monitor (VGA, this was before HDMI was common on laptops) just like the Vaio, as well. Oh, and it had built-in wireless and ethernet, just like an ultrabook (does the Air have ethernet anymore, or would htat make it too thick?). Oh, and a mic and headphone port. And a webcam. And a card reader. And a 10"-class screen. And a too-small keyboard. Am I leaving anything out?

Re:SSD failure rates

[BronsCon]

Certainly. When you've got as many platters as I do spinning around in an enclosed space, you'd best be conscious of cooling and power stability.

Re:SSD failure rates

[hobarrera]

Netbooks are small, low performance devices.
Ultrabooks are lightwight high performance devices.

How is an ultrabook an expensive netbook? They're on the opposite ends of the performance charts.

Also, I still see netbooks everywhere in the market for a very very low price.

edit

[djupedal]

errr
1.5% of a 4TB SSD that sells for USD$29,000 is roughly 60 GB = $425.

Hard drives warranty

[danbob999]

5 years should be mandatory by law. If you can't support your drive for 5 years, you shouldn't be allowed to manufacture hard drives at all.
I don't understand this new trend in making new hard drives with only 1-2 years warranty. The same goes for SSD.

Re: Hard drives warranty

[Anonymous Coward]

"Shouldn't be allowed"? I hope you are joking.

Re:Hard drives warranty

[TheGratefulNet]

margins are paper thin. no time to do QA. what ends up happening is that we, the buyers, are the 'remote QA dept'.

sad but true. we have to test the hell out of things we buy for the first 30 days.

profit profit profit! isn't extreme capitalism wonderful? sigh ;(

Re:Hard drives warranty

[mysidia]

I don't understand this new trend in making new hard drives with only 1-2 years warranty. The same goes for SSD.

If it shaves cost off the unit; there are people who will buy it, and take the chance.

I would say that the manufacturers have a right to offer them this option.

In fact; I would say manufacturers have a right to provide options with less than a 1 year warranty.

Re:Hard drives warranty

[Burz]

In the 2000s consumers became almost the exact opposite re: warranties as they were in the late 80s/ early 90s when a good warranty seemed to matter as much as any other criteria. I've been trying to buck that trend, but until the last couple years it was almost impossible. When I shop for electronics that have no moving parts and are *not* portable, the warranty has be be at least 3 years and this even includes some moving-parts items like hard drives. My two most recent HDD purchases (and some that I helped friends and clients with) had 5 year warranties.

The thing about insisting on a 'long' warranty is that the price then becomes an aid in finding equipment that is actually more reliable. Among stable brands, the cheaper models in the longer warranty class will tend to be more reliable; A higher confidence level from the manufacturer is often reflected in the lower price. Likewise, the junkier models will get higher price tags in order to be able to cover the higher failure rate. Nowhere is this more obvious than with computers that have options to purchase mfg extended warranties.

Of course, even if the prices are the same, getting equipment with a higher failure rate is still a raw deal because of the cost of downtime, possible data loss, shipping, etc.

Re:Hard drives warranty

[citizenr]

margins are paper thin.

WD and Seagate have a healthy 40% margin on every drive they sell.

Re:Hard drives warranty

[tftp]

margins are paper thin. no time to do QA.

Not just margins. Development time is short. A model of the drive has to be produced and sold in less than a year, and replaced with a new model after that. Who can afford an endurance test, even if accelerated?

Re:Hard drives warranty

[Anonymous Coward]

Manufacturers have no rights. They're not people.

Re:Hard drives warranty

[MikeBabcock]

Except they've priced their drives so low that you're looking at 40% of very little... on a per-unit basis, they're still making very little money and better have a very low return rate to account for it.

Re:Hard drives warranty

[citizenr]

You are not making any sense. Im sure $4B a year net profit is scraping the bottom of the barrel in your book.

Re:Hard drives warranty

[Anonymous Coward]

Then people have the right to purchase a hard drive with less than a 1 year warranty.

Re:Hard drives warranty

[the_humeister]

Depends on revenue. If revenue is $5 billion, then no. If revenue is $200 billion, then yes that is scraping the bottom of the barrel. BTW, Western Digitals net profits were just under $1 billion. That's about a 6.6% profit margin. It's respectable, but nowhere near, for example, Apple's 26% or Intel's 20%.

Re:Hard drives warranty

[hedwards]

I've found that with HDDs it makes more sense to get the cheapest one that fits the need I have as I can and just go in for a proper UPS. I don't think I've had a single HDD fail to make it at least 5 years since I started that policy. Google some years back was looking into the reliability rates on HDDs and their basic conclusion was that there isn't any meaningful difference in terms of the expected lifespan of a HDD.

That was some years back, it would be interesting to see if that's still the case with regular HDD. But, with backup services so cheap, I'm not as worried about the occasional HDD failure as I used to be. Especially now that I regularly check for bit rot and can restore anything that goes bad.

Re: Hard drives warranty

[Tablizer]

America = freedom to sell crap

We fight in crappy wars for the freedom to make and sell crap. Men and women died in these crappy wars so that you can purchase all the crap you want.

Re:Hard drives warranty

[xlsior]

I don't understand this new trend in making new hard drives with only 1-2 years warranty. The same goes for SSD.

It's very simple, really: Because they can.

The main reason is that there's only three hard drive manufacturers left in the world: Seagate, Western Digital, and Toshiba. (Samsung & Hitachi's HDD divisions have both been aquired in recent years, although you can still find drives with their brandname on them, for now)

Out of those three, only WD and Seagate manufacture large capacity 3.5" HDDs. It's essentially a duopoly.
When there's just two players left that are both manufacturing at pretty much full capacity, there's very little incentive left to offer long warranties -- that just costs them money in the long run. Warranties have been trending downwards, and it's unlikely that'll change any time soon.

Re:Hard drives warranty

[Anonymous Coward]

That your comment's modded +5 Insightful shows how little Americans understand freedom and choice.

Re:Hard drives warranty

[countach74]

Manufacturers are not people, but people do run the manufacturing companies and as such those people (especially the owners of said companies) do have the right to do business as they please. If they make enough poor choices and end up alienating their market, they will likely go out of business. Meanwhile, another manufacturer will eventually come onto the scene to offer people what they want. Business cycles: they are a good thing. Requiring companies to provide {x} length warranty is ridiculous, as a business does tend to try to satisfy their customers; if they choose not to provide such a warranty freely, it is probably because it is not important enough to their end users.

Remember, when you pass legislation saying {x} type of product must have {y} or {z}, all you effectively do is increase the price of that product and when prices go up, demand goes down. In other words: you likely won't be as interested in the product if you have to spend another ${n} on it. :) My point is, you can't have your cake and eat it too. You can't have the same product at the same price point and a longer warranty. Either the price must go up or the quality down in order to compensate for additional cost to the manufacturer.

Re:Hard drives warranty

[thegarbz]

5 years should be mandatory by law.

Justify your choice of warranty periods. The vast majority of consumer electronics have warranty periods of 1-2 years. Why should a device with highly sensitive mechanical components with tolerances that make you wonder why the hell it ever worked to begin with be subject to some tougher warranty standards?

I'd greatly prefer my $1k TV have the longer warranty period than a storage device costing about as much as dinner for 2 in town.

Re:Hard drives warranty

[stiggle]

Some other countries (eg. most of Europe) have a "fit for purpose" condition in consumer law.
What this means is if you sell a high end product then it is expected to last longer than a cheap product. So if you bought a high spec top end laptop then its expected to last longer than a no-brand $150 one. So even if the warranty only says 1 year - it can be argued that its "not fit for purpose" and you can get it replaced/fixed/refunded.

eg. Cellphone with 1 year warranty but sold on a 2 year contract. The carrier expects the phone to last 2 years as thats the period of contract they sold with the phone, so if the phone fails in that time then you can claim it was not fit for purpose.

Re:Hard drives warranty

[bill_mcgonigle]

I don't understand this new trend in making new hard drives with only 1-2 years warranty. The same goes for SSD.

Most of my hard drives have died either very quickly or after about 3 years. I would count on one replacement during the 5-year warranty. So, when they cut warranties to 1 year, it at least doubled the cost of hard drives.

Not sure where that plugs into the inflation calculator...

Re:Hard drives warranty

[Zeromous]

Wisdom brother.

  without properly scaled ups and diesel generator I experienced drive failures of about 1-3 per month accross 300 machines. 3 power failures where ups gave out (which may have only directly caused 5% of those failures)

After moving to a properly resourced DC...
3 per year.

Re:Hard drives warranty

[Anonymous Coward]

So you want to required that commercally available technology be two years behind the rest of the world? They do a good bit of testing before setting the warranty.

Re:Hard drives warranty

[Anonymous Coward]

Really, more laws? I am not an extreme libertarian by any standard but this is a bit much. Why not let the market sort it out (as it is)? If you mandate a 5 year warranty, I can guess what the unintended consequence will be: prices jumping back to $2/GB or even higher.

Re:Hard drives warranty

[Threni]

> eg. Cellphone with 1 year warranty but sold on a 2 year contract. The carrier
> expects the phone to last 2 years as thats the period of contract they sold with the
> phone, so if the phone fails in that time then you can claim it was not fit for purpose.

Nice in theory. I'm in the UK and my HTC Desire broke after little over a year, in an 18 month contract. Orange refused to pay the £60 it cost HTC to repair (it was a faulty usb socket). I left Orange 6 months later and don't buy HTC stuff any more. I suppose I could have taken time off work and gone to court. There's a market here, I support, for people to do this sort of thing on people's behalf.

Re:Hard drives warranty

[danbob999]

Because if any other part of my PC breaks I can replace it easily. However if the hard drive breaks, I loose all my data. I don't care about the drive itself but it sucks.
We require higher quality drives. A few years ago, hard drives were cheap and had 5 years warranties. Now prices have gone up by 50% and warranty is now 1-2 years.

Re:Hard drives warranty

[Anonymous Coward]

If it shaves cost off the unit; there are people who will buy it, and take the chance.

I would say that the manufacturers have a right to offer them this option.

In fact; I would say manufacturers have a right to provide options with less than a 1 year warranty.

Yes, but manufacturers shouldn't have the right to bury us in garbage. It may be acceptable, so long as they are responsible for properly separating and recycling the parts (and not just dumping them in China). With that condition though, I doubt that 1 year warranties would still look as attractive. Even so 1 year is far too short for virtually anything, excepting toothbrush heads and rechargeable batteries.

Something really needs to be done to discourage the mass production of disposable crap. Even major appliances fall into that category these days, and that is just wrong.

Re:Hard drives warranty

[Anonymous Coward]

5 years should be mandatory by law. If you can't support your drive for 5 years, you shouldn't be allowed to manufacture hard drives at all.
I don't understand this new trend in making new hard drives with only 1-2 years warranty. The same goes for SSD.

What if I want to save the money on a warranty because I buy bigger hard drives every year or two anyway? It that such shitty behavior that my government should be forcing me to subsidize your consumer preference?

If it is, I guess my momma taught me manners wrong.

Re:Hard drives warranty

[Anonymous Coward]

No they don't. They'll purchase what the manufacturers allow them to purchase.

Re:Hard drives warranty

[AmiMoJo]

In fact; I would say manufacturers have a right to provide options with less than a 1 year warranty.

I think most people would prefer to know that anything they buy has a minimum 2 year warranty (Europe), rather than having to check every single god-damned thing. The freedom to get fucked over by manufacturers isn't freedom at all, it's just a waste of your valuable time and money.

Re:Hard drives warranty

[Anonymous Coward]

Yet another in there "Must be a law for everything" camp.
How about a law for disposable pens must last 5 years or you aren't allowed to make them.
When will we learn that all of lives problems arnt solved by passing more laws.

Re:Hard drives warranty

[Kjella]

The real reason is that I do care about durability, but I don't particularly care much about the warranty. If a drive with a one year warranty craps out after three years, I have to buy a new one. If a drive with five year warranty craps out after three years, I don't have to buy one but I still have to RMA it, ship it in, lose whatever I didn't have backup of (which of course never happens), I'm down on capacity until the new one arrives and then I have to install it and restore from backup. Or you have a degraded array that may or may not rebuild with a spare, if you swing that way. Of course you pay a premium on those five year warranty disks which is like an insurance to cover replacements, but are you actually getting any better results? It's certainly a non-zero risk in either case, so you must have the same backup regime anyway it's only a matter of how often it happens.

Re:Hard drives warranty

[Anonymous Coward]

In other words: you likely won't be as interested in the product if you have to spend another ${n} on it.

Not necessarily. If a drive with a 1-year warranty costs x amount of money, and a drive with a 5-year warranty is x+n (where x >> n), and either drive can be expected to last between 1 and 5 years (which seems reasonable based upon past experience), then I'll likely spend less in the long run if I buy the drive with the five year warranty.

Re:Hard drives warranty

[MikeBabcock]

Some people don't do math well so I just ignore them but good reply.

Re:Hard drives warranty

[Twinbee]

How can Apple be only 26% profit margin when they price their stuff at least double that of other manufacturers?

Re:Hard drives warranty

[just_common_sense]

Make sure you have redundancy and/or backup so that you don't lose your data. Your hard drive will eventually die, so if it lives a long time that simply means that you're going to lose more stuff when it eventually does bite the rust.

Re:Hard drives warranty

[danbob999]

Yeah but 1 year warranties are a sign that the manufacturer do not trust its own drive.
RAM manufactuers have lifetime warranty. Why? Because those who build it trust it. Also because they know that you are not going to RMA your faulty 64 MB 133 MHz SDRAM from the past millenium.

Re:Hard drives warranty

[thegarbz]

Yes but I want a pony. The point is the same. You're holding the only component with critical moving parts who already are messing with the laws of physics to a higher standard than the rest of the entire electronics industry.

Sorry but if you're worried at any time about dataloss you're doing it wrong.

Also the change in warranties hasn't necessarily reflected the change in drives, just a nickel and dimming from the industry. I don't think there's anything to say that drives are more likely to fail now than they used to be despite the warranty period.

Re:Hard drives warranty

[mysidia]

I think most people would prefer to know that anything they buy has a minimum 2 year warranty (Europe), rather than having to check every single god-damned thing.

The people who don't should nevertheless have the right to buy, and not have to pay for such a long period of manufacture insurance against defects; just because most people may be loss averse, and like to spend the extra money for extra warranty, that they probably still won't get to use; does not mean manufacturers should be required to force this financially irresponsible choice on all their customers (the manufacturer has to add to the price of the product MORE than the expected average cost of this warranty, when extending it from say 30 days to 2 years: therefore, the longer warranty will never be in the customer's best interests -- unless they already knew in advance they'd get one of the defective units, OR the manufacturer was fiscally irresponsible and did not increase the product cost appropriately to cover the warranty costs PLUS risk premium).

Re:Hard drives warranty

[danbob999]

With a shorter warranty, the industry has no insensitive in building drives lasting longer. As long as they fail after the first year, they are fine. It's now a duopoly so you can't really vote with your wallet.

Re:Hard drives warranty

[thegarbz]

You mean we get to hurt it's feelings? Oh right incentive, sorry but that's still false. If what you said were true the premium electronics industry wouldn't exist.

Look at your typical point and shoot camera. Many come with 1-2 years warranty. Most are plastic. Most aren't expected to last beyond 3-4 years tops. Now look at the top of the line DSLRs. They also come with 1-2 years warranty yet are built like a brick shithouse often capable of many hundreds of thousands of shutter actuations.

The incentive is simple as IBM showed the world 12 years back. When your customers start nicknaming your flagship product after it's failure rate your business effectively ends, and it did, and they sold their HDD division to Hitachi for a song after the Deskstar (Deathstar) debacle.

And of course the opposite is also true as can be seen with the car industry. Hyundai and Mitsubishi made awful cars. Horrendous even by the low bar set by Australian car manufacturers. When both companies effectively lost much of the market in Australia their attempt at re-igniting the brand came with declaring a 10 year warranty on all cars. It didn't work. Mitsubishi has effectively pulled out of the Australian market. Hyundai nearly pulled out and were only saved by a dramatic increase in quality and drop in price over the last 5 years to bring their cars in line with consumer expectation (oh and they have since dropped their 10 year warranties too).

Re:Hard drives warranty

[proibido]

Agree with you in almost every argument, except:

1) when the manufacturers agree price and conditions between them, consumers get screwed;

2) Consumers get screwed during the process of the manufacturers "natural selection" as you put it, there must be some reasonable way to protect them from the low quality / charlatans / evil manufacturers;

But but but but

[Anonymous Coward]

I thought "solid state" was supposed to be so reliable? And why can't we 3D print SSDs? Hm?

Re:But but but but

[Zynder]

Fail Troll is FAIL

And after the warranty expires?

[Anonymous Coward]

"Yet, according to IHS, client SSD annual failure rates under warranty tend to be around 1.5%, while HDDs are near 5%."

And outside of warranty...? The great majority of individuals will still be using their SSDs long after the warranty expires. This is even more true for businesses. A better measure of drive reliability would be one taken over an extended period of time, one that more accurately mirrors typical drive usage.

how long's the warranty?

[Laxori666]

That under warranty less SSDs fail doesn't mean they outlast HDDs... If warranty is 1 year, and all SSDs fail in 1.5 years, yet hard drives usually fail only in 3 years, hard drives are still better off.

In other news, Laxori666 was too lazy to RTFA and is hoping someone will chime in. He is tired and drowsy and so he will blame it on that when in fact, he would have done the same regardless - except perhaps without this addendum as such honesty usually requires some sort of altered state of consciousness.

Re:how long's the warranty?

[GigaplexNZ]

Warranties are typically longer on SSDs than HDDs these days. I can't comment on the specific models they used in the article however.

What a pun...

[Anonymous Coward]

Ba-doomp-CRASH!

SSD failure rates

[Anonymous Coward]

What is it that is said about lies and statistics ?

if I harddrive survives the 100 hours it is likely to last a long time, an ssd on the other hand will die.

I have lots of experience with spinning disks, not so much with ssd, but what I have had is mostly bad, be it firmware bugs that wipe all the data or just outright failures. I have one today in a laptop where the drive was toasted but still passed the bios diagnostics which were built into the laptop.

lease them

[Neo-Rio-101]

Yet another reason to never buy Mac.

You're absolutely right.

That's why we lease them for work;

Bit error rates are more important to monitor

[JoeyRox]

While catastrophic drive failures make headlines what's more likely to happen during the useful service life of both HDDs and SSDs are unrecoverable media/bit errors and these may ruin your day as much as a catastrophic error. If you look at the bit error rate of any contemporary HDD and compare it to its capacity you'll come to a startling conclusion - an unrecoverable read error is rated to occur once every 2 to 5 times the full capacity of the drive is read. SSDs have about the same unrecoverable read error rate.

Re:Bit error rates are more important to monitor

[Dorianny]

In hdd's a mounting number of bit errors or frequent controller resets are red flags pointing to a imminent drive failure. In my experience ssd typically fail catastrophically without any warning signs at all, something which is pretty rare in hdd's.

Re:Bit error rates are more important to monitor

[Anonymous Coward]

please mod this guy up - that is exactly the problem with SSD. Not that fail less or more than HDD but that whne they fail its lights out and with no warning. HDD usually afford some warning and even after death data are often partially or fully recoverable.

Paucity of information....

[QuietLagoon]

The author of the summary and/or TFA seems to draw a conclusion based upon a paucity of information.

Yet, according to IHS, client SSD annual failure rates under warranty tend to be around 1.5%, while HDDs are near 5%. So an SSDs not only outperforms, but on average outlast spinning disk."

The unknown in the equation is the length of the warranty periods for the drives used in the comparison.

Re:Paucity of information....

[wisnoskij]

And also what happens afterwards. Most drives will not fail under warranty, most should fail many many years afterwards.
This does not tell us if the average SDD fails 1 week after its 2 year warranties runs out, or if the HDD lasts for 8 year longer on average.

Re:Paucity of information....

[dfghjk]

But why would you assume they are different?

"Most drives" will not fail under warranty but so what? It's not informative.

Drives have early life failures. Once that period is past they will typically last a long time. That's common for a lot of things.

What evidence do you have that SSD and HDD are fundamentally different in this regard? Lacking that, your comment is worthless.

Re:Paucity of information....

[thegarbz]

This is not necessarily relevant. Size, speed and power consumption advances have meant that I only actually have 1 harddisk out of 6 computers and a file server that is currently not in its warranty period.

Data size is increasing and putting strain on older storage. When I can buy a 2TB harddisk for $100 why should I care if my old 250GB is under warranty? It's essentially disposable.

Re:Paucity of information....

[thegarbz]

The unknown in the equation is the length of the warranty periods for the drives used in the comparison.

Don't assume that is necessarily a negative thing for the SSD. Harddisk manfuacturers dropped their warranties on consumer drives from 5 to 3 years last year. The cheapest SSD I have seen has a 3 year warranty and the majority have 5 years. I wouldn't be surprised if the statistics in this case are biased against the SSD.

Re:Paucity of information....

[Anonymous Coward]

I have no evidence of that.
But neither is there any evidence that they are the same.
So really they are just making assumptions and drawing useless conclusions from those.

Re:Paucity of information....

[Anonymous Coward]

Which, barring evidence to the contrary, is *more* sound than assuming they behave *differently*, and drawing useless conclusions from there.

my hard drive is going bad

[Anonymous Coward]

my hard drive has developed bad sectors on the platter. SMART says there are some reallocated sectors, pending pending sectors and uncorrectable sectors. the hard drive fails all of Seatools's tests. oh yeah, the bad sectors are at the end of the Master File table of the NTFS. Maybe i should buy an SSD now instead of a computer. lol

3 weeks

[Anonymous Coward]

I got my 1st ssd this spring, it was awesome and fast, for 3 weeks.

Re:3 weeks

[camperdave]

I got my 1st ssd this spring, it was awesome and fast, for 3 weeks.

What a coinicidence... so was mine. Still is, as a matter of fact. Oh, and so is the one in my laptop.

Re:3 weeks

[PPH]

I've got an EeePC with an SSD. Bought back in 2008. Still runs fine.

Apples and oranges

[Anonymous Coward]

If an SSD fails after its shorter warranty has expired, does that mean it's more reliable than a traditional hard drive that's lasted longer, but is still under warranty? No? Then what's the point of comparing failure rates "under warranty" if the warranties can be unequal?

To make SSD's look good, I imagine.

Oh, and also traditional hard drives often make alarming noises several hours to days before they fail, giving you time to back them up. SSD's, not so much.

Re:Apples and oranges

[Your.Master]

I think it's a good question whether warranties are unequal.

I think it's a bad assumption that they are slanted to SSDs (for instance, the demand seems more likely to be there for SSD warranties given the reputation). It might be true but it's not obviously true.

Re:Apples and oranges

[GigaplexNZ]

Hard drives typically have shorter warranties than SSDs. OCZ has a bad reputation for having reliability issues compared to other SSDs yet their warranties are 3-5 years. The typical Seagate and WD warranties are now 2 years.

Yawn.

[Anonymous Coward]

Anyone who isnt using a SSD by now for at least their boot drive is stuck in the past.
It's the single best upgrade you can make anymore.

Either way stop the fucking articles about it.
Leave them with their warm feelings for spinning rust full of multi gigs of stuff they never touch.

They'll wise up eventually. Or not.
Either way it won't hurt you any. Enjoy your speedy pc and laugh at the rusties if you must.

Re:Yawn.

[Anonymous Coward]

Not sure if troll or what. It takes less time to boot and log in than it takes me to go for a cuppa, I only have to do it once and the rest of the day it is cached. And that is if I shutdown the pc, instead of just sending it to sleep.

A SSD for a pc is cute tho.

Re:Yawn.

[dbIII]

Anyone who isnt using a SSD by now for at least their boot drive is stuck in the past

You've just summed up those stupid applications on MS Windows with hard coded paths to "C:" drive. They still exist.

Re:Yawn.

[blackraven14250]

You can use Windows' version of symbolic links to make that work across multiple drives.

Re:Yawn.

[Billly Gates]

I still use mechanical drives as SSDs are shaky quality wise and only have 1 million writes. I pound my disks with hundreds of gigs of VMs.

SSD is very expensive for non casual users and the reliability with the limited writes are not there yet. I know I may get modded down by some fanboys here who swear by it, but I see risk for the sake of quicker boot times.

I/O is not much better if at all than a mechanical drive. It is latency with lots of small files which makes SSD 100x faster but to me this just means booting faster and some apps may shave off a few seconds. I save $$$$ with using what already works with my 1 TB drive.

Re:Yawn.

[0123456]

Anyone who isnt using a SSD by now for at least their boot drive is stuck in the past.

I boot my work PC about every two months.

It's the single best upgrade you can make anymore.

If you spend all day just booting your PC. Otherwise, a faster CPU or GPU or more RAM is likely to be far more useful.

Re:Yawn.

[dbIII]

I've never heard of that actually working in such situations. Do you have a link to any example or even anything that points in such a direction?

Re:Yawn.

[Anonymous Coward]

I wonder why booting is the ultimate test of your pc? Maybe if thats all you ever did.

On ssd every program starts faster. reads all data faster. writes all data faster.

If you actually DO anything other than start firefox once a day or something. It's a HUGE speedup. Huge.
Going from 50MB/sec on a good day to several hundred MB/sec all the time.

What used to take minutes now takes seconds. Or less.

I can't even imagine why you wouldn't want that. You'll spend hundreds on a small % gain in cpu speed. But won't spend hundreds for a 1-2-3-400% gain in data i/o.

It's one of the last major bottlenecks in a modern pc. The drive system. And you like it being slow?

Is there an onion on your belt?

Re:Yawn.

[0123456]

I wonder why booting is the ultimate test of your pc?

'Cause, uh, the OP said 'Anyone who isnt using a SSD by now for at least their boot drive is stuck in the past.'

See. BOOT DISK. The disk you BOOT from.

If you actually DO anything other than start firefox once a day or something. It's a HUGE speedup. Huge.

I got 32GB of RAM. After a few days, anything worth running is in the disk cache for the next few months until it's rebooted.

Seriously, buy a real PC and you won't make silly comments about 'an SSD is the best upgrade EVAH!' any more. It was a great upgrade for a our netbook, because we're often booting it to check a web page and shut it down. It's not a great upgrade for a PC with lots of RAM that's booted every few months.

Re:Yawn.

[gilboad]

I'll give you the benefit of doubt that you're not simply trolling.

In the last two years I've experience two SSD bricks on my main Xeon workstation (2 x X5680, 36GB RAM, 6+1 x 320GB enterprise SATA in software RAID6, running Fedora 19 x86_64).
On the other hand, the 5 (!) year old 320GB enterprise SATA drivers are working like new (hence I've yet to replace them).
Now, back when I had the SSD's I used them as a fast cache, but for the life of me I couldn't feel the difference. (Can same the same about the occasional breaking).
Sure, firefox would launch *marginally* faster, but:
1. I boot once every major kernel release (or major security issue).
2. With anywhere between 10-20GB of free RAM (depending on the number of active VMs) most of the software I used is cached.
3. Even when compiling a large project, CPU is usually the limiting factor (even w/ 48 jobs).

So, would I feel the difference on a single disk laptop? Sure!.
Do I feel the urge to add a third SSD to my workstation? Do I trust them enough with my work? Doubt it.
Guess I'll have to see how the Linux kernel's bcache works and how well it handles bricked SSD's.

- Gilboa

Re:Yawn.

[Archimonde]

SSDs have their place even if you don't want to acknowlede it. On my workplace we've upgraded 4 year old laptops with SSDs because those 60+ laptops were unusable how slow they were. An you know what happened after upgrade? Everything is flying fast on those laptops. That was much better upgrade than boosting ram from 4 to 8 GB (max supported) of RAM (we've tried that one, difference was miniscule).

It just doesn't matter if the user is booting, rebooting, opening outlook, browsing, or whatever, evething is reasonably fast.

And don't get me started on how my retina macbook pro with 16GB of RAM and a SSD is fast. It is ridiculously fast in day to day work. I absolutely don't have to wait for a computer to do something. When you get used to this kind of speed and responsiveness, when you get back to a relatively fast computer (good cpu, board, ram etc) with a normal HDD the difference is astonishing as everything seems to be dog slow. And the other poster is right traditional HDD is the biggest bottleneck on the whole system bay faaaar.

Re:Yawn.

[Anonymous Coward]

Yeah, Windows "symlinks" are not at all like Unix symlinks. Unix symlinks work in every case all the time, they look just the same as the path they're pointing to. Windows "symlinks" actually almost never work. I've tried to use them to solve problems, but they only seem to work reliably in the Windows file explorer and in a couple other cases, most software that accesses files just doesn't follow or even see the symlinks. They're useless. And yes, that was actual NTFS "symlinks" I tried, not the cheap .ini hack.

Captcha: unparsed

Re:Yawn.

[Anonymous Coward]

Anyone who isnt using a SSD by now for at least their boot drive is stuck in the past.
It's the single best upgrade you can make anymore.

The only thing that a SSD does for the common user is, at best, shave a couple of seconds at boot time and start unstarted apps which have a heavy initialization process a tiny bit quicker.

And for those overwelming benefits one needs to cough up around 200 dollars.

For the common user, that's not an upgrade. That's wasted money.

Re:Yawn.

[Anonymous Coward]

No.

Your system isn't kept in ram at all times. System files will be loaded when you use the computer.
If you bought you computer some time during the last decade a CPU upgrade is probably the most useless upgrade ever. Seriously, better heatsinks/fans might improve performance more.
GPU is pretty much always a nice upgrade.

It is interesting that you mention RAM. When you buy more RAM it is usually not because you need more things active in memory but to compensate for HDDs being completely fucking worthless when it comes to read/write access and is generally only good at storage.
What most people have realized is that adding extra RAM for disk buffers only really help with write performance. Once you get over 4GB adding extra RAM usually don't help at all.
When you computer has at at least 4GB RAM and a reasonable GPU an SSD should be the next upgrade. Not for the storage but because more RAM can't compensate for the suckiness of HDDs.

Re:Yawn.

[Anonymous Coward]

"Anyone who isnt using a SSD by now for at least their boot drive is stuck in the past."

Twaddle. Add the rest of your post and it's just pretentious twaddle. And aluminum doesn't rust.

Re:Yawn.

[Anonymous Coward]

Even worse. Put your games and the large work-programs on a SSD and the PC will feel light it runs at warp 5. It really is no comparision.

If you still have 3,5" drives kick them out (SSD or 2,5") or your PC will never be silent.

Re:Yawn.

[Anonymous Coward]

And pray that some program that you use will not mess them up.

Re:Yawn.

[Anonymous Coward]

Anyone who isnt using a SSD by now for at least their boot drive is stuck in the past.
It's the single best upgrade you can make anymore.

FUCK YOU. I'll continue to live in the past so long as the past tech works better. I have 34TB available to me on my main computer (on which I am typing). Shove your fast read slow write small drive up your arse. I'll find something else to do while it boots.

Re:Yawn.

[newcastlejon]

I used symlinks (or whatever the Windows name is) before Steam introduced multiple libraries so I could have some games installed on the SSD and others on an HDD. There's a great shell extension here that lets you create them. Since then I've reformatted and so I no longer need them for Steam, but they do come in handy for splitting my home folder across drives. The tool I've linked to also does junctions but I've never had to use those.

Re:Yawn.

[Anonymous Coward]

You're just a home user, aren't you? It's easy to tell -- a professional wouldn't dare write off what is still often the right tool for the job.

Re:Yawn.

[Dr.Dubious DDQ]

I'm not sure finding a way to shoehorn a relatively expensive SSD into my laptop to decrease my boot time from 10-15 seconds down to 5-10 seconds is really going to be worth the cost and effort, personally.

Re:Yawn.

[frinkster]

Anyone who isnt using a SSD by now for at least their boot drive is stuck in the past.

My spinning platter drive allows me to spend far more time in the present.

Re:Yawn.

[swb]

+1 on this.

A $100 spent on a decent (eg, Samsung) SSD takes a low-end Core2 system and makes it really usable again.

I have salvaged nearly a dozen client systems with SSD upgrades -- 3-4 year old Core2 systems that went from intolerable to downright speedy with a drive swap.

99% of the time RAM isn't enough -- you can't stuff enough RAM into the box, and even if you do, Windows RAM caching just isn't sufficient for the virtual memory and DLL thrashing you get with a dozen app windows and web pages open. Video card upgrades do nothing for generic desktop users. Gigabit NICs have been built-in forever, so that isn't an option nor would it do anything for most users anyway.

Re:Yawn.

[Anonymous Coward]

> SSD by now for at least their boot drive

I upgraded the C: drive on my home Windows system and one at work, and I can't tell a difference in performance except when doing an update from Subversion. If you run Windows, there's no real reason to have an SSD. It just isn't fast enough to take advantage of it.

On my late-2008 MacBook, upgrading to an SSD was like getting a new computer. I was ready to sell it before the upgrade, but fourteen months later, I still have the laptop. The SSD added more than a year to the laptop's life. It was worth every penny.

Re:Yawn.

[UneducatedSixpack]

So all you do is just boot the laptop and that is it? When using windows SSD is major improvement for systems that have to boot frequently. At work I do not have SSD and I do not need it because I just reboot once every few weeks/months and after a while everything becomes cached in the RAM. At home everything is quite different. I boot my computer couple of times a day and usually either check email, browse web, run game, do some development or photo stuff. In this case SSD is really great as I can boot computer in about 10 seconds and then start anything I want in 1-2 seconds. With HDD everything is much slower in this scenario.

Re:Yawn.

[blackraven14250]

It actually depends on what exactly the behavior of the program is, but in many cases you can use either a symlink, hard link or an NTFS junction to make it work.

Re:Yawn.

[WuphonsReach]

Otherwise, a faster CPU or GPU or more RAM is likely to be far more useful.

That depends on how often you need to do two (or more) things at once on the computer, both of which touch the hard drive.

If you are a heavy multi-tasker, constantly opening up new programs, opening up new files, scanning directories for changes (version control), the CPU/GPU/RAM only takes you so far.

Once you have enough RAM to keep your normal working set in memory, additional RAM doesn't add much speed at all (except use as a bigger drive cache). The GPU only helps if what you are doing can be offloaded to the GPU in the first place. A faster CPU only helps, up to the point that you are waiting on that spinning platter of rust.

Having spent almost 3 years using a SSD on my laptop (I jumped in once prices dropped below $2.50/GB), I can't stand desktops that use 7200 RPM drives (and heaven forbid it is a 5400 RPM or flex-speed drive). Trying to start up something like Firefox, Thunderbird, or other applications takes 5-10 seconds or longer. And if you try to do anything else with the machine at the same time (like open up an office application or check on something), you are just prolonging the agony.

IOPS matters, even for the lowly end-user desktop/laptop. With the SSD, you no longer have to take a 5 minute coffee break while you open up all of your applications.

Re:Yawn.

[WuphonsReach]

+1

I'm still using my Thinkpad from 2007 (an early Core Duo model). It has 8GB of RAM, Win7 and a SSD in it. While the Core Duo is showing its age (and I'd love 6 or 8 cores and 32GB of RAM), I can't complain much. I was ready to replace it 3 years ago because the old HD was so slow that it bothered me constantly.

Our office desktops are AMD64 x2 machines from 2006-2009. We're in the process of upgrading them to Win7, 4GB+ of RAM and 128GB SSDs. At which point, they'll be good to go for another 6-7 years.

Cheap upgrade at this point and it makes the machine almost vanish from user-perception.

Re:Yawn.

[toddestan]

Well, all I can say is that I jumped on board early with SSDs. After nothing but problems I went back to 'spinning rust' on my desktop PC. Why? Because it works. The marginal speed increases after the PC has booted aren't worth the wasted time and headaches of using a technology that, for whatever reason, doesn't seem to have matured yet.

Re:Yawn.

[Anonymous Coward]

You sir, are an asshole in penis clothing.

Stay away from OCZ and SandForce

[JDG1980]

OCZ's failure rates are higher than the rest of the industry's by an order of magnitude. Also, earlier SandForce drives have reliability problems because the firmware was written by paranoid loons who were deathly afraid of reverse-engineering and the drive goes into irrecoverable 'panic mode' when any abnormality of any kind is sensed. I think that newer SandForces (post-LSI acquisition), especially Intel's, are less likely to do this, but the original failures still taint the brand with the stigma of flakiness.

If you stick with Samsung, Intel, and SanDisk, you should be fine. Stay away from OCZ at all costs, and be skeptical of any SandForce drive not made by Intel.

Re:Stay away from OCZ and SandForce

[Anonymous Coward]

OCZ's failure rates are higher than the rest of the industry's by an order of magnitude.

I see this said often enough that it's probably true in general.

However, I've had two OCZ Vertex drives since 2009 and a Vertex 3 since early 2012 that still work fine, so I must have done something wrong.

Re:Stay away from OCZ and SandForce

[Anonymous Coward]

1) Those are pre-25nm and 2) Yes, you're just plain lucky not to suffer a controller failure/firmware bug with any OCZ drive.

Re:Stay away from OCZ and SandForce

[k8to]

*slightly* bigger datapoint. Work bought around 10 ocz drives for shared build machines. 3 failed within the first 6 months. This was around early 2012.

I wasn't really paying attention to the choices being made around then or would have warned away from that choice. I only found out after I started doing my own research into card-based flash storage that I asked about what we used (the second pass was violin cards) and found out we'd used ocz drives for the first failed experiement.

Re:Stay away from OCZ and SandForce

[Anonymous Coward]

Pretty much this. I had an OCZ Vertex 2 bite me from the shutdown/power on bug one in a 256 chance every boot of striking it, extremely piss poor on OCZ's part especially the bit where the drive panics then no access is available surely read only would have been the sensible approach or failing that provide a tool to fix the drives themselves afterall it wasn't really "hardware" at fault.

Re:Stay away from OCZ and SandForce

[DigiShaman]

I've had a bad rash of Vertex 2 and 3 version drives in the past. Forums filled with problems of the drives "bricking". There was a period when Newegg offered a sweet deal on a Vertex 4. I purchased it and soon after a firmware update was released. The industry revisited the drive and benchmarked the Vertex4 one last time after the firmware update. It got faverable reviews and the price jumped up on Newegg. So far, I've now got a total of two Vertex 4 drives in use in both my MacBook Pro and desktop at home. No problems and the speeds have been consistent.

Stay away from OCZ? I bit harsh advice now days. I'm not saying you should give them equal treatment along with the three brands you've mentioned, but OCZ isn't some shitty low-end no-name brand either. Give them another chance should you run into another price drop of their stuff.

Re:Stay away from OCZ and SandForce

[PhunkySchtuff]

I can't verify the reasons you've given, but I can back up those failure rates.
When SSDs were still crazy expensive, OCZ were at the more affordable end of the range. It got to be that they failed so much whenever I put in an RMA for one with my supplier they wouldn't even ask for details or attempt to troubleshoot to verify the fault.
Me: "Hi, I've got an SSD for a warranty return"
Them: "OK, have you got the serial number or the original invoice?"
Me: "It's an OCZ"
Them: "Oh, no worries, we'll courier a new one out to you"

Since prices have dropped, I now only use Intel and have had a grand total of one failure, in a 4 year old 80GB disk.

Re:Stay away from OCZ and SandForce

[Miamicanes]

> but OCZ isn't some shitty low-end no-name brand

You're right, because a shitty low-end no-name brand at least has the benefit of anonymity instead of a name that's tainted beyond repair or redemption.

Re:Stay away from OCZ and SandForce

[DirePickle]

OCZ's SandForce firmware was junky for a while and caused all sorts of problems, but AFAIK that was sorted out quite a while ago. I bought a Vertex 3 a couple of years ago, and had some issues for the first week--updated the firmware to the latest and have not had an issue related to it yet. Aside from that, I don't know that they have any actual hardware-related reliability issues that don't plague all of the manufacturers.

Re:Stay away from OCZ and SandForce

[Anonymous Coward]

You taint the brand with your comments about shit that happened years ago.

I wonder what grudge you hold or how much they are paying you.

Re:Stay away from OCZ and SandForce

[Jesus_666]

What about Samsung? I've heard good things about their SSDs - but then again I only briefly surveyed the market before deciding that getting an SSD big enough for my needs would be prohibitively expensive.

On the other hand, if they build SSDs like they build smartphones I wouldn't want to touch one...

Re:Stay away from OCZ and SandForce

[Jesus_666]

To answer myself: Yes, Samsung is good, which the very post I replied to says. My reading comprehension is godlike.

Re:Stay away from OCZ and SandForce

[Anonymous Coward]

AData is using also SandForce controllers, I got two of them, works perfect, also they were the first to enable TRIM

http://goo.gl/2QWewq

Re:Stay away from OCZ and SandForce

[Raleel]

I manage a small cluster of 8 machines with 24 OCZ vertex 3s a piece.

My own experience was that all failures for these were based on bad firmware, either on the raid controller or on the disk themselves. We had massive failures for a while there... 1 a week, 2 a week...

then we upgraded the firmwares on both and it all just went away. We've lost 1 disk in 2 years since.

Re:Stay away from OCZ and SandForce

[trparky]

I've had good luck with two of the Samsung 840 Series SSDs that I have and my brother has had good luck with his 840 Series SSD as well. I've had better luck with the two Samsung SSDs that I had with an Intel SSD. Funny, the Samsung SSDs were cheaper than the Intel one I had but the Intel one died. Oh well.

Some people say that the reliability of the Samsung SSDs come from the fact that Samsung made the thing, the whole thing. Not just a piece of it. Samsung made the NAND, the main PCB, the controller chip, and the controller chip firmware. The whole thing was developed and manufactured in-house unlike several other SSDs that use the Sandforce SSD controller which has historically been quite a buggy piece of shit.

Maybe things have changed with the Sandforce controller but after that one Intel SSD that had the Sandforce controller in it that died on me, I won't trust a Sandforce controller-based SSD again.

Re:Stay away from OCZ and SandForce

[Anonymous Coward]

Early Kingston drives were horrible as well. I bought a whole bunch of them to replace mechanical drives on laptops at work and 2-3 years later, about half of them have stopped working. I think it's the controller, since you can't even read drive info. Meanwhile, Intel and Samsung drives have worked flawlessly so far.

Re:Stay away from OCZ and SandForce

[toddestan]

Hey, at work we had an IBM Deathstar last in a computer that was powered near 24/7 until sometime around 2010 or so. Every once and a while, they'll be a fluke or two.

Re:Stay away from OCZ and SandForce

[thexile]

Vertex 4 utilized OCZ own Everest 2 controller, not SandForce.

Bad statistics?

[MouseTheLuckyDog]

What are the warantee periods? Are SSDs shorter. Is it the same usage scenario -- for example using the SSD for swap?

What do the failure curves look like?
I suspect for HD's it's a gaussian and for SSD's it's a skew normal distrubution with the scew leaning towards the end. Meaning a large amount of SSD fail past a certain. time while many HD still work? Have we even seen enough SSD's in the wild to see that failure yet?

Re:Bad statistics?

[gagol]

you need swap?

Are they including D.O.A. ?

[dbIII]

I've heard several people state that they've bought SSD drives that just would not work when they got them home and they had to do an exchange. Do these statistics include those returns or only ones that failed in service?

Re:Are they including D.O.A. ?

[wisnoskij]

SDDs and HDDs. There is a huge percentage that arrive dead, or die within a week. But I do not think SDDs fair any worse that HDDs.

Re:Are they including D.O.A. ?

[MikeBabcock]

All electronics fail on a bathtub curve ... they either fail predominantly when new, or when old, and rarely in the middle. That's why burn-ins are common among the knowledgeable -- it gets you past that initial failure stage before you've used the drive for something important.

Re:Are they including D.O.A. ?

[dbIII]

Either way anecdotally the SSDs are turning up DOA a lot and the HDDs are not. Perhaps it comes down to a lack of testing for one and testing for the other?

Re:Are they including D.O.A. ?

[Miamicanes]

Actually, the big problem with SSDs is the fact that unlike conventional hard drives, SSDs often DO have catastrophic data loss out of the blue long after the burn-in period, but long before the drive is even remotely "old". A conventional hard drive that's going to die tends to do it when the drive is new, before you've invested much in setting up Windows yet again. A Sandforce-based SSD is most likely to die 3-6 months after installation, then every 3-6 months after each "secure erase" resurrection, until the day you put it into panic mode trying desperately to recover a file you saved an hour ago that didn't make it to your latest backup.

Re:Are they including D.O.A. ?

[Anonymous Coward]

No, it comes down to people *noticing* that the 'new tech' stuff dies, but being *used to it* when the 'old tech' stuff dies. You read a lot about how SSDs die, so you assume they die more often, so it sticks in memory longer than what you assume is a 'rare' instance when the traditional 'spinning disk' HDD dies.

It's an example of confirmation bias. (As demonstrated by the fact that 5% of HDDs fail during their 1-3 year warranties, while only 1.5% of SDDs fail during their 3-5 year warranties.)

Then again, you're also largely looking at anecdotes here from people who are willing to draw sweeping conclusions from statistically insignificant sample sizes (less than 10 drives *total* over the course of a decade).

Re:Are they including D.O.A. ?

[MikeBabcock]

Actual numbers from actual studies please.

Anecdotes are not data.

Re:Are they including D.O.A. ?

[dbIII]

My post was all about questioning those statistics - hence the subject heading. An immediate return and exchange may be catagorised differently to a warranty return as it is with some other products, so I'm suspecting that may be the case, but as I don't know I've asked the question.
What I do know is some models of OCZ drives a while back with a specific controller had a very high rate of immediate returns, well into double digits, which led to OCZ using a different controller. Since there were a lot of that model surely it would skew the total if immediate returns are included?

Major fault. Linus lost an ssd

[chromaexcursion]

another story on /.
a TRULY dead ssd is impacting the linux kernel release.
one in Linus's server.
bad timing, to try to pump bad statistics.
there are lies
damn lies
then there are statics


better to go with the lies ...

and hire better tech aware ad men

in my experience...

[Anonymous Coward]

... a significant percentage of HDD failures occur within the first month of life, resulting in very little actual data loss, as people still retain prior backups.

Does anyone have hard numbers for the DOA rates of each type of drive?

Ability to recover

[ArchieBunker]

Now for the useful information. How many of the failed SSD's were they able to recover data? I suspect not many.

Re:Ability to recover

[Anonymous Coward]

Who cares? That's why you use them for your boot drive and not to store your porn collection.

Re:Ability to recover

[Barlo_Mung_42]

Why would anyone ever need to try and recover data from a dead drive when they can just get it from the recent backup? Oh, you thought you could just push everything to the internet and let us back it up for you? Hmm, gosh, maybe that's not a very good plan.

Re:Ability to recover

[Barlo_Mung_42]

Funny thing is, the porn collection is one thing we can trust the internet to back up for us.

Re:Ability to recover

[thegarbz]

Now for the useful information. How many of the failed SSD's were they able to recover data? I suspect not many.

Spoken like someone who doesn't have a very useful backup strategy.

If my SSD fails it's going in the bin. My likely losses extend to the cache in the browser session from this Slashdot visit. You should not be hoping to recover data from ANY failed drive.

Re:Ability to recover

[Anonymous Coward]

Nope, only if you're into vanilla porn... Niche porn is quite hard to find twice, once the links are dead the odds are low to none. Except of course for the boring old crap that always gets reposted everywhere.

Captcha: lasers. Woohoo!

Re:MicroSD cards = great SneakerNet

[noh8rz10]

why use a balloon?

Recoverable Failure rate: 99.9% HDD, 1% SSD

[citizenr]

Wrong stat.

Yes, things to break, but its important HOW they break. HDDs have very 'nice' failure modes. You can recover bits from the platters as long as you do not put one in MRI machine or a fire. SSDs just DISAPPEAR from the system with data and encryption keys to that data and NO ONE including manufacturer can do recovery (they can put flash chips in reader and read encrypted bytes, but encryption keys were in the controller that just died).

How about another one: Warning before failure rate? Again 90% HDD, 1% SSD.

Do you know how many SSDs survive running out of spare sectors? Again about 1% :) 99% just die without going into read only mode.

Re:Recoverable Failure rate: 99.9% HDD, 1% SSD

[GigaplexNZ]

There's a reason people back up data rather than relying on data recovery on a broken drive.

Re:Recoverable Failure rate: 99.9% HDD, 1% SSD

[0123456]

There's a reason people back up data rather than relying on data recovery on a broken drive.

You must know different people to the ones I know.

Re:Recoverable Failure rate: 99.9% HDD, 1% SSD

[gagol]

I have AAA, why bother with the engine check lights, right? Who needs it?

Re:Recoverable Failure rate: 99.9% HDD, 1% SSD

[GigaplexNZ]

If you're using engine check lights as an analogy to SMART data so you know when to swap out a drive, I still disagree. Once SMART data goes bad, you can usually still get data off the drive but I wouldn't trust it to not have some corruption. I'd just yank the drive, rebuild the RAID if the machine needs high availability, or copy over from backup if it doesn't.

If recovering from backup and the backup isn't quite up to date I'd probably try some kind of data recovery, but I wouldn't want to rely on it.

Re:Recoverable Failure rate: 99.9% HDD, 1% SSD

[Miamicanes]

Actually, there's an even bigger question... was the "failure" definition for SSDs carefully-crafted to exclude drives that experienced total data loss, but could technically be "repaired" and run for another 3-6 months before the next instance of total data loss?

At least when a traditional hard drive dies, it stays dead instead of coming back like an unstoppable undead zombie with an insatiable appetite eager to make a second attempt at destroying whatever data of yours it missed the first time around.

I can't think of a better gift for that special "frienemy" in your life than a Sandforce SSD. It's a gift that just keeps on giving... misery. Endless, eternal misery.

Christmas wishlist: solid state storage that rocks

[WaffleMonster]

Lack of a good replacement for spinning platters in 2013 is a little depressing.

SSDs consume nearly as much power as spinning disks, writes are destructive high current cap banks faillure prone average I/O per day lifetime ratings assume just *minutes* at max write performance per day, bit density/cost sucks and lots of free space required for effective wear leveling.

Personally I'm sticking with spinning platters until memristers or something replaces SSDs. While ago rumor was it would be late this year or next year before product starts shipping. I can wait/save up.

With regards to the failure rates according to TFA most models of Western digitals provide the same ~1.5% figure.. I don't buy from other vendors or have the outlier WD model so their contributions don't effect me personally.

Also take percautions in assuring proper airflow/temperature range, not spinning down, vibration dampening mounts and never shipping ground when ordering online.

So the non-failing hard drives

[tompaulco]

So, if 5% of hard drives are failing in the first year of warranty, then the other ones have to last 180 years on average in order to meet the MTBF specifications of 1.5 million hours that hard drive makers claims. Because surely they wouldn't lie to us.

Re:So the non-failing hard drives

[GigaplexNZ]

They only need to run 1.5 million drives for 1 hour to reach that MTBF. If 5% of them die after 2 hours they still hit the MTBF.

Re:So the non-failing hard drives

[bytesex]

Are you not confusing 'mean' and 'average'?

It takes a ream but...

[Anonymous Coward]

Paper only fails when water is involved.

Infant mortality vs wear

[mlwmohawk]

Statistics are wonderful things, if you choose the right one you can make any case you want. I want to know more about the warrentees. I want to hear about the nature of the issues. Recoverable errors vs complete death. Infant mortality vs just wear.

Re:Infant mortality vs wear

[gagol]

I agree, all HDD that failed me gave me some audio warning for like a week before ultimate failure.

My own experience

[TheGoodNamesWereGone]

Last year I bought a used 90GB Kingston V200 used off Craigslist. I am running my Linux installation right now off of it. At first I treated it with kid gloves and tweaked everything I knew how to tweak to reduce writes, including turned off swapping entirely-- with 16GB RAM I never touched the swapfile anyway-- reduced the swappiness system variable, etc. When I went from Mint 13 to 15 I only turned off the swapfile and it's still running like a champ, but a part of me still just doesn't trust the little darling. I decided to get a larger one to run my W7 installation on and bought an Intel 520-series 180GB one. It too has been rock solid reliable. I trust Intel more than Kingston and probably will take W7 off it and put on Linux the next time I upgrade my OS. Anyway, the gist of what I'm saying is, they've both done very well for me. I have no personal experience with OCZ but everything I've read says stay away. Samsung makes the drives that go in Macbooks, though I don't know if the 840 you can buy at retail is the exact same one as in the Apples. For anyone thinking of taking the SSD plunge, I say go ahead. I highly recommend Intel, but then I really don't boot to Windows that often these days so the the 520 gets *much* less use than my little Kingston. It's been a workhorse.

Re:My own experience

[DigiShaman]

Apple has crippled their OS to *only* enable TRIM on Apple branded SSD drives. Otherwise, if the drive is flagged anything but Apple, TRIM is disabled even if the drive supports it. This has been true going as far back as Snow Leopard. However, you can still enable TRIM on non-Apple SSD drive, but it enables a perlscript hack to a file that ignores the vendor flag check. The script to use depends on the version of OSX you're using however. Oh, and you must apply the hack after each service pack update (normal updates seem to be fine however) as it will replace the modification back to normal operation.

not as good as it sounds....

[smash]

... if i need to store x TB of data, i can set up a RAID1 mirror and have fault tolerance for way less than the cost of SSD based storage to hold it in a non-fault tolerant manner.

Re:MicroSD cards = great SneakerNet

[Anonymous Coward]

well, condoms also work, but I prefer bareback.

Re:Put in perspective...

[Anonymous Coward]

Industry damage control

importance of stored data

[Hanzie]

Who cares? That's why you use them for your boot drive and not to store your porn collection.

^^ +1 insightful, for having correct priorities. I wish I had mod points today.

Probably just me...

[symbolset]

Visualizing Linus, on an old laptop in a hotel somewhere busily merging the kernel of the world's most popular OS. Probably in his jammies. How the world has turned.

5 percent is too high.

[edibobb]

I question the 5 percent annual failure rate for hard drives. I haven't had 1 in 20 hard drives fail ever. That's at least a couple of magnitudes off the MTBF numbers.

Re:5 percent is too high.

[rubycodez]

any single person's experience isn't relevant

Just lost my first SSD

[swilly]

Talk about timing. I'm right now recovering data from my first SSD failure (an almost three year old OCZ Vertex 2). As failures go, this couldn't have gone better. I'm able to read the drive, but I can't write to it. I wish all drive failures were this nice. I'm having Newegg overnight me a Samsung 480GB SSD as a replacement. I should probably think about replacing the two SSDs that are older than the one that failed, just in case.

Just this year I've lost two 1TB hard drives, and one of them somehow corrupted my (thankfully backed up) RAID 5 making it unrecoverable. So, I decided to replace the older consumer grade 1TB drives with 3TB WD Red drives (supposedly enterprise grade), and what do you know? One of them is dead on arrival. WD replaced it with a "recertified" drive, which is annoying, but at least it works.

I also lost a Blu-ray drive, so it hasn't been a good year for my storage devices, but so far my anecdotal experience has SSDs with better reliability than mechanical drives. YMMV.

Re:Just lost my first SSD

[yuhong]

Yea, flash have limited write cycles, so this kind of failure can happen. What would be nice is if RAID controllers were aware of it and were able to rebuild RAID arrays by copying data.

Different types of failure...

[viperidaenz]

There's DOA and there's failed after use.

I'd be willing to bet a fair amount of the hard drives are DOA. You don't lose data if you never put any on it.

SSDs fail

[l3v1]

What I'd like to know, why the hell can't they make SSDs that have a life in the ballpark of a ram module. Yeah, you wouldn't buy new ones so often, but come on, crap is crap. Oh, now I can hear all the tablet and ultrabook fans (I also use ultrabooks with SSDs for quite a while) that they are good and nice. But it's not just those devices that SSDs are used in. E.g. we have an application that needs lots of ram and also needs lots of disk space with quick access so we use SSDs for caching important parts. And we've seen SSDs die in weeks (!), while we have 1-2 failed ram modules per year. Yes, different tech, still: make higher quality crap, that's all.

SSDs should still be handled with care.

[gilboad]

Semi-OT: A word of friendly warning:
A couple of months ago (year?) I bricked a 120GB Intel 520 w/ the latest firmware (not sure
if it was 400i) w/ ext4 on Fedora x86_64. (Second bricked SSD in 12
months)
A *very* short power shortage crept under my APC UPS and bricked the SSD.
Amazingly enough, the power shortage didn't crash the machine - which
continued working off the main HDD software RAID array.
Luckily for me I rather distrust SSDs (see below) and use it as fast
cache-of-sort, so I only lost a couple of hours of work. (If any)

IMHO SSDs have one huge drawback: Unlike HDDs that can be partially
recovered from more-or-less any type of damage by recovering data
around bad sectors or replacing a fried controller board, SSDs complex
write scheme and the resulting complex firmware usually means that any type of
damage / firmware error will completely bricks it leaving more or less zero
chance of getting the data back.
On the top of that, we (as in all of us) have 40+ years worth of
experience in predicting the life cycle (and death) of HDDs. There's
far less information about the life cycle of SSDs.

Case of point: A couple of days after this incident a family member lost one of his HDDs.
Unlike my dead SSDs, with some work I managed to recover 95% (or more) of his files.

Don't get me wrong. SDDs will replace HDDs in the end - but in the mean time, I'd keep SSDs for non-critical tasks.

- Gilboa

Re:SSDs should still be handled with care.

[Miamicanes]

You forgot to mention the wonderful way Sandforce controllers encrypt the data with a key that you (the drive's owner) aren't allowed to have, so your ONLY data recovery option (on the rare occasions when it MIGHT be an option) is to pay an extortionate amount of money to one of Sandforce's "trusted partners" to decrypt it for you... and apparently, they actually charge more money to do what's now a 100% automated software-based recovery than the same companies USED to charge to remove the platters from a conventional hard drive in a cleanroom and mount them in a recovery unit.

I'm frankly surprised that some company like OCZ hasn't come up with an "innovative" new-economy (as in, "fuck you, consumer!") business model for SSDs, like selling 256-gig drives for $25 that lock themselves after some random period of time between 180 and 720 days after first use, and charge $2,000 for the key to unlock the drive (decreasing to $1,000 after a week, $500 after a month, $250 after 6 months, $125 after a year, and $64 after 2 years). Is there anybody who doubts that netbook manufacturers would pee their pants with glee if they could take advantage of that kind of "innovation"?

Ancient data.

[Reeses]

All this discussion on this and no one has commented that TFA is from 2011??

This article isn't reliable information. It's from when SSDs were relatively new and definitely doesn't apply to the in-the-field results people are seeing in 2013.

Re:Ancient data.

[Anonymous Coward]

That's really some kind of cosmic coincidence, because in 2010 I bought a 2TB eSATA HDD for my DirecTV HD-DVR. That damn DVR keeps the drive running and ticking 24/7. So for 3 years now that $130 HDD has been running non-stop with zero issues.

Boot from RAID 1 SSDs?

[Futurepower(R)]

What about putting 2 SSDs into a software RAID 1 configuration? Does that solve the problem?

What you said is my experience, also. I haven't had catastrophic failure of a HDD in perhaps 20 years in a population of perhaps 15 computers. In my experience what most often fails is the HDD electronics, so it is possible to extract the data by temporarily replacing the HDD electronics with a circuit board from another, identical HDD. Also, of course, in the last 20 years we have replaced HDDs because of frequently replacing computers.

Re:Boot from RAID 1 SSDs?

[jkflying]

That won't work if they both die from some bug which is triggered by eg a certain write sequence followed by TRIM, then power cut in the middle of TRIM. They will both be killed.

Re:Boot from RAID 1 SSDs?

[jcdr]

That won't work if they both die from some bug which is triggered by eg a certain write sequence followed by TRIM, then power cut in the middle of TRIM. They will both be killed.

Ok, but is this kind of systematic bug only theoretical or some finding tend to prove that there are a reality ?

 

Re:Boot from RAID 1 SSDs?

[mlts]

This may be a waste of engineering, but if a HDD is enterprise grade with a high MTBF, I wonder how difficult it would be to make a RAID 1 controller that would do loosely coupled writes, and keep a ring buffer in battery backed up RAM (so if power goes out, the in-flight writes to the slower drive would be stored.) That way, writes to the SSD would complete, and with enough RAM, the writes to the HDD can complete when it gets around to it. Of course, when the RAM is full, I/O gets blocked until the buffer is dealt with so the drives may be out of sync time-wise, but are always consistent with each other.

Re:Boot from RAID 1 SSDs?

[bugs2squash]

I'm surprised that the SSDs are not internally organized as multiple units and then use some built in raid controller. I would have thought that the user could then choose how to organize the drive for maximum IOPS, capacity, write endurance, or reliability even on a machine like a laptop that has no space for multiple drives.

Re:Boot from RAID 1 SSDs?

[mlts]

That would be nice, perhaps separating metadata into multiple, redundant areas. No metadata, no translation tables, and the drive has no clue what logical sector is stored on what set of physical cells.

I think a lot of SSD makers are going for capacity/price, and the failure rate being close to HDD is "good enough", although done right, it should be a lot lower, just due to the fact that there are no moving parts that can wear out. I'd definitely pay more for drives which address the weakest points (power going off unexpectedly) of SSDs, and some type of pre-fail system that warns, then sets marginal sectors read-only as opposed to just dropping the drive off the face of the earth when the wear limit is reached.

Re:Boot from RAID 1 SSDs?

[Anonymous Coward]

I'm surprised that the SSDs are not internally organized as multiple units and then use some built in raid controller.

But they are arranged like that. The SSD controller doesn't physically perform reads and writes, it has a digital interface to the flash memory and it sends commands (read / erase / write) to individual flash die. (In fact, it's common for each die to be subdivided into two "planes", which are independent and can process commands concurrently.) A reasonably large SSD often has 64 or more flash die, and relies quite heavily on keeping lots of commands in flight to sustain high IOPs, sequential throughput, and so on.

I would have thought that the user could then choose how to organize the drive for maximum IOPS, capacity, write endurance, or reliability even on a machine like a laptop that has no space for multiple drives.

You can optimize just about any SSD for better performance and write endurance simply by choosing to leave some of the drive empty when you partition it. Just make sure to perform an ATA Secure Erase (or TRIM the whole drive) using the manufacturer's tool before you partition, so that the empty space is truly "free" as far as the SSD is concerned.

Re:Boot from RAID 1 SSDs?

[stoatwblr]

Some devices ARE arranged internally as RAIN (Redundant array of Inexpensive NAND), but most manufacturers just pack as much in as they can and use sparing to achieve reliability.

Unsurprisingly, like powerfail protection (PFP), It's more common at the top end/enterprise end than consumer drives.

As time goes by you'll see RAIN + Sparing + PFP + WriteVerify become standard on all but the cheap'n'nasties.

USENIX study

[Anonymous Coward]

https://www.usenix.org/conference/fast13/understanding-robustness-ssds-under-power-fault

No, they can't

[dutchwhizzman]

They sell several amounts of already soldered chips on the main board. That means that you will have to either replace components on the main board and reprogram the SSD controller, or replace the entire main board.

Re:No, they can't

[Guy Harris]

They sell several amounts of already soldered chips on the main board.

Not soldered to the motherboard for the 15" Retina MBP and not soldered to the motherboard for the 13" Retina MBP. On which Macs is the SSD soldered to the motherboard?

Re:No, they can't

[Anonymous Coward]

Go to the Apple Store, configure a 15" Retina MBP and find this greeting:

Storage
Your MacBook Pro comes standard with 256GB or 512GB of flash storage. Please note that the flash storage is built into the computer, so if you think you may need more storage capacity in the future, it is important to upgrade at the time of purchase.

Re:No, they can't

[Anonymous Coward]

The original Air had the chips soldered IIRC.

Re:No, they can't

[Guy Harris]

Go to the Apple Store, configure a 15" Retina MBP and find this greeting:

Storage Your MacBook Pro comes standard with 256GB or 512GB of flash storage. Please note that the flash storage is built into the computer, so if you think you may need more storage capacity in the future, it is important to upgrade at the time of purchase.

What that means is that Apple won't upgrade it after you buy the machine. "Built into the computer" does not, as the pictures on the iFixit pages indicate, mean "soldered onto the motherboard".

Storage

[Anonymous Coward]

I cannot think of one storage medium that doesn't ultimately fail after enough time, except possibly stone tablets stored in a cave.

My own experience

[Anonymous Coward]

I won't say my experience mirror that of the worldwide average ...

I do not have the make/model of ALL the SSD that we use, nor I have the range of make/model of the harddisk that we use either.

From the computers (from server to desktop to laptop to tablets to specialized devices like xerox machines) that my company uses, total number in the thousands, SSD failure is on par with that of HD --- and their average lifespan is actually shorter than their harddisk counterpart.

Corsair?

[antdude]

What about Corsair Force Series F115 (115 GB; CSSD-F115GB2-BRKT-A)? :P

Cry me a river

[Kuruk]

Cry me a river

Great USENIX study on SSD under power fault

[advid.net]

Out of 15 SSD tested, only 2 are failure proof under power fault (only one maker and model).
(yes, I've read the pdf)

I'd like to know who is the winner, the anonymous vendor/model called "A-2".
It is not the most expensive, almost the cheapest, but it has at least a power-loss protection.
Another vendor has power-loss protection but his models failed the tests.

Direct link to pdf and figures erratum.

Bit Corruption: SSD#11, SSD#12, SSD#15
Flying Writes: none
ShornWrites: SSD#5, SSD#14, SSD#15
UnserializableWrites: SSD#2, SSD#4, SSD#7, SSD#8, SSD#9, SSD#11, SSD#12, SSD#13, HDD#1
Metadata Corruption: SSD#3
Dead Device: SSD#1
No failure: SSD#6, SSD#10, HDD#2

Their last word conclusion :

We recommend system builders either not use SSDs for important information that needs to be durable or that they test their actual SSD models carefully under actual power failures beforehand. Failure to do so risks massive data loss.

Thanks again for this link to the Usenix study, too bad you posted anon (patent need mod up).

shrinking geometries reduce reliability

[lkcl]

i was speaking to someone who works in aerospace: they have deep concerns about the geometry shrinks in the chase for extra storage. the smaller the geometry gets, the less reliable it gets, it's as simple as that. they are having enormous difficulty getting hold of large-geometry small-capacity NAND flash ICs.

also, i've begun to replicate the drive-torturing software which was mentioned a few months ago here on slashdot. one SSD i tested which is reported to have good power-loss protection failed in THREE minutes. another took 24 hours and 2,500 power-cycles.

Nonsense

[Kleokat]

I have seen SSDs fail too often to use them for anything but read-only storage.
I have gone back to HDDs simply because they are more reliable and lasts 10 times longer.
An SSD in a desktop computer lasts for max 2 years, and if you power it off while it is working, then it is a sure way to kill the entire drive or at least all the data on it.

SSDs are great toys, but never use them for anything other than read-only data.

Completely believable

[Murdoch5]

I've been using SSD's for a few years now and I'm NEVER going back to traditional hard disks for boot drives. I moved myself over to SSD boot disk a few years ago, maybe 5, In the last 5 years I've had 1 SSD die and it was replaced with no questions asked. If I look at how many traditional hard disks I lost in 5 years when I was using them as boot and storage disks I would probably average 2 or 3 a year. SSD technology is quickly becoming the only sensible solution for computing, not only does it last N times longer then traditional hard disks, it's faster and more rugged.

I go for speed + LONGEVITY

[Anonymous Coward]

Dr. Dobbs Journal anyone? This quote's WHY I steer clear of FLASH SSD's:

"SSDs do have one important flaw, however, which is that the cells that store the data can be written to only a finite number of times before they can no longer be used reliably. Most SSDs solve this problem by distributing writes all over the disk, so that no one group of cells gets rewritten too frequently. How much of a problem this will be is hard to tell. I've spoken with some users who have had to replace SSDs due to this issue. If consumer devices see only four or five years of light use, I expect most of them will work fine for the expected life of the device. Power users, such as developers, however, may encounter this problem more frequently." - From http://www.drdobbs.com/architecture-and-design/the-demise-of-hard-disks/240161048

As that's something I'd DEFINITELY run into eventually, since much of the work I do & have done, is File I/O oriented (by the 100's - 1,000's quite often) - & it's the "WHY" of why I don't go FLASH based SSD's... yet (until they are not only mature, but PROVEN, for the reasons stated above - perfornance not an issue anymore, but rather durability!).

---

So - Do I use SSD's here? Yes! However, NOT Flash RAM based hardware units (but rather those based on DDR-2 RAM or PCI-2.2 SDRAM):

( & for a LOT LONGER than most worldwide have typically!)

Since 1992 or so, 1st using separate HDDs (slower seek/access by FAR) & then using software ramdisks per the list below (on a MS-DDK based one I wrote in fact, on how I apply them):

Then applying Software-Based Ramdrives to database work with EEC Systems/SuperSpeed.com on paid contract (which did me VERY WELL @ both Windows IT Pro magazine in reviews, & also MS TechEd 2000-2002 in its hardest category: SQLServer Performance Enhancement & SuperSpeed.com too - since I improved their wares efficacy by up to 40% via programmatic control & tuning programs for them) - which, only the past few years now it seems, OTHERS are finally "latching onto" for performance purposes in database work in industrial environs! The EEC/SuperSpeed.com unit had 1 great thing going for it - mirroring back to HDD to save state of data!)

I move the following off my wd Velociraptor SATA II 10,000 rpm 16mb buffered harddisks that are driven off a Promise Ex-8350 128mb ECC ram caching raid sata 1/2 controller (which defers/delays writes via said cache, & also lessens physical head movement on disks & this is where I am going to make it even faster via lessening its workloads, read on & reduces fragmentation as well in the same stroke - "bonus") onto my 4gb DDR2 Gigabyte IRAM PCIExpress ramdisk card 2006-present (& before it, a CENATEK "RocketDrive" 4gb PC-133 SDRAM based one on PCI 2.2 bus circa 2002-2006):

---

A.) Pagefile.sys (partition #1 1gb size, rest is on 3gb partition next - this I didn't do on software ramdrives though)
B.) OS & App level logging (EventLogs + App Logging)
C.) WebBrowser caches, histories, sessions & browsers too
D.) Print Spooling
E.) %Temp% ops (OS & user level temp ops environmental variable values alterations)
F.) %Tmp% ops (OS & user level temp ops environmental variable values alterations)
G.) %Comspec% (command interpreter location, cmd.exe in this case, & in DOS/Win9x years before, command.com also)
H.) Lastly - I also place my custom hosts file onto it, via redirecting where it's referenced by the OS, here in the registry (for performance AND security):

HKLM\system\CurrentControlSet\services\Tcpip\Parameters

(Specifically altering the "DataBasePath" parameter there which also acts more-or-less, like a *NIX shadow password system also!)

---

* All of which lessen the amount of work my "main" OS & programs slower mechanical hard disks have to do, "speeding the

The name's school, old school.

[bjoswald]

I don't know if mechanical HD's are considered "Amish technology" in the PC world yet, but I've been using computers since the 80s and have yet to have an HD disaster. Could be luck, could be karma, but I have no desire to switch yet. Even the typical sales pitch of "bigger, faster!" doesn't sell me like it does with GPUs.

Mature VS New tech

[DarthVain]

OK LOL!

HDD give you plenty of warning now. In fact most of SMART tech, and a host of other things to run continuous tests looking for potential failure, as well as OS that specifically look for indications as well. Now.

Years ago, this was not the case. You MIGHT get some warning depending on how it decided to fail (bad sectors etc.,,), however most back in the day gave you about one second of actually notice before dying in a grinding crunching sound, or in a small black puff of smoke. I suppose in that light, you aren't wondering what the matter is, as you know it died, as it had the good measure to give a last death rattle before departing into the dark night.

Backups had to be done all the time, because at any time, it could go poof. Now you get like a weeks warning and can go pick up an external drive at your leisure (which is what I did the last time I had a HD fail).

The reason we have the protections is because they were so bad, and consumers demanded better drives, driven by consumers. SSD drives have only been mainstream commercial for a handful of years. It is pretty new technology compared to HDD. Give them a second to catch up!

I know mine is never going to be replaced.

[DarthVain]

As I got a mSATA SSD which plugs into the back of my motherboard, and my case has no removable back plate.

I hope that thing never fails, as if it ever does there is no way in hell I am going to disassemble everything just to get that POS out.

That said, if it is still under warranty I will probably have to, sigh. Word to the wise, if your building a system using a mSATA SSD, you might want to make sure your case has a removable back plate.

Though the thought of using a dremel to cut a hole just briefly passed through my brain, though I suppose the MB wouldn't react too kindly the the pile of steel shavings it would produce...

Note to Fractal Designs: Your Node 304 uses ITX format motherboards, most of which support mSATA SDD, and there is limited HDD space in the case. How much more would it really cost you to make a removable back plate? Not to mention the back plate that most modern heat sinks now require also (and yes I forgot this and had to redo).

MLCs vs shrinks vs NANDs vs.......

[unixisc]

The MLC aspect is the biggest problem, even more than the SLC shrinks. Having 4 different well regulated states in which to store data, even while the core VDD keeps decreasing due to the shrinks, makes data corruption even easier. If I know that an SSD is based on MLC technology, I'd avoid it like the plague.

As the shrinks take place, it happens for not only NAND flash, but NOR flash as well. As NAND flash hits the several TB range, we know that NOR flash can as well. If they make SSDs based on SLC NOR flash, and price them somewhat higher than the current SSDs, they can actually realize the claimed theoretical advantages of flash memory over HDDs, since w/ NOR flash (think of the flash memory used in the BIOS of your PC motherboards), you don't have the issue of corrupted bits, low data retention or any of those associated issues

It's a costs issue for most people

[p51d007]

I'm going to be rebuilding my 4 year old computer later this fall. I am going to see if I can install a smaller SSD for booting, and a huge spinning drive for applications/storage. The price is still too high to get a large enough hard drive to really store anything. In a practical standpoint, I think off storage, which is what I do for a lot of my photos, is better than leaving them on a drive. DVD/external drives for critical stuff, and internal for day to day usage is what I do, so an SSD for boot/working programs would be all I would need. My current HDD's on my computer, have pretty much been running 24/7, with the occasional off period when a storm hits, power goes off and the UPS kills the computer.

More marketing spin eh?

[cronos1013]

Yea, I have to call bullshit on this one. I have been building PC's for years, since the end of the 1980's. On only 3 or 4 occasions since I have started building machines have I had spinning platter drives die on me while I am still trying to use them (typically 2-3 year replacement cycle for me). Since I have been purchasing SSD's, I have lost one a year, consistently, because the NAND fails. I have bought SSD drives manufactured by Corsair, WD, OCZ, and most recently a Samsung. Only the Samsung is still alive, but it's only 2 months old. All of the others stopped working at right around a year, you get a replacement under warranty, and then that one dies within a year. These drives have gone into both my machines and those of my kids. These drives are heavily used, there are a bunch of read/write, but with these new claims of the drives lasting a "normal user" like 70 years, it's bullshit for them to die for ANY user within a year consistently. These claims appearing on tech blogs is complete marketing spin and a disservice to normal users, as only ONE of the drives made it more than 12 months, my OCZ, which lasted 14. I live in an air conditioned condo with no water problems, normal humidity, and the drives are positioned so as to allow proper cooling (with a fan blowing over them). It isn't an environmental issue killing them, it's the fact that these things aren't made to last. I still have had the same Caviar Green platter drive for data for 4 years now, and while it isn't as fast as an SSD, it hasn't died either. Until the in practice life expectancy gets up for these thins, users who want better speed will be forced to continually replace them year after year (I now use a mirrored setup to prevent data loss). The performance is great, but it's a shame that the longevity is awful.

Fewer fail under warranty...

[Anonymous Coward]

But the warranty for SSD's is shorter than the equivalent sized HD's Also, SSD's are susceptible to something platters aren't: Being entirely wiped by a buildup of static electricity in the computer. I work in a PC repair shop, and no less than 10 times a week, I see static buildup. On average I work on 6 machines a day.

Yes, Brand matters, but then some people compared country shop, and ferrari, the thing is, my Kia Soul will spend a lot less time in the shop than a ferrari will, and the ferrari costs much more than 10x the amount of the Kia. More expensive does not equal higher quality.

The average life span of a platter based drive that is > 1TB is less than a year, from what we've seen in our shop. The average life span of a platter based drive between .5 TB and 1TB is 1 to 3 years.

The average life span of a platter based drive less than .5TB is 3 to 5 years.

The average life span of an SDD drives that we have seen, (ALL less than 1TB) is .5 yrs to 1.5 years.

Calculated Failure rates, be damned, the reality of the probability if you buy more than 1 drive from the same manufacturer, from the same retailer, is that the drives will be from the same lot, with the same manufacturing flaws. Chances of multiple drive failure goes up exponentially.

We've seen it happen w/ all the mfg. So the theory goes that the more drives in the array, the lower the chance of complete failure, but the reality is usually the opposite.

I will say this, most of the arguments I've seen posted here on both sides are based on Theory, and or reports from places that have highly specialized operating environments that will do much to limit the failure rate of any computing device. But throw the same equiment into real world application, i.e. not running under ideal circumstances, then of course the failure rates go up, and are expected by those in the real world, and many times go unreported, because it's considered part of life. Just because a tree fell, and no one was there to see it, doesn't mean it didn't happen.

Not really being fair to the guy, IMO ....

[King_TJ]

For starters, SSDs are being marketed as mainstream replacement for HDDs -- not as exotic high-performance upgrades. (So that Ferrari analogy is pretty much invalid here.)

Sure, they outperform regular HDDs -- but Core i5 and i7 CPUs outperform older Pentium 4 core-duos too, and those outperform PIII processors by a long-shot -- yet you don't expect them to be far less reliable just because they're faster. If anything, the promise was the lack of mechanical moving parts would virtually guarantee they last longer.

The problem is HOW they fail...

[Anonymous Coward]

I don't trust those 1.5 vs. 5% numbers. It fails to distinguish between types of failures. And the article is talking exclusively about Intel SSDs (often SLC ones at that), not the whole market.

HDDs usually fail by developing increasing numbers of bad sectors. While some abruptly drop dead, that's quite rare. The result is that you tend to get some warning, and the catastrophic data loss rate is quite low.

SSDs have a nasty habit of bricking themselves, as Linus experienced. And InterServer reported at the end of page 4 of the Tom's Hardware article.

Crucial

[Anonymous Coward]

Before I went SSD, I read through the user reviews.

With hundreds of people chiming in, you can get a pretty good spread of information to sample.

I also found a spreadsheet helpfully provided by an IT guy working a data center with thousands of drives under industrial load. He listed the up times and failure rates by vendor.

Nobody makes failure proof devices, but some are a *lot* better than others. I couldn't afford the top of the line, but Crucial back around 2011 was doing really well and was quite reasonably priced. Their M4 series seemed to have a lot of happy customers. So I now run a few of those. They haven't given me any trouble.

I don't know what things are like now, (this article is a couple of years old, btw), but I imagine I'd run a similar product sample when I need to buy new drives in the future.

It's all done in metadata so not like a file

[dbIII]

The sort of programs we are discussing don't know enough about NTFS to be able to deal with the difference between the links and real files. They are not presented to the application as if they are real files as in other filesystems with symbolic links - it all depends on the application understanding extra metadata and if the writer of the application has not thought of that it doesn't happen.
Thus often pointless for the sort of programs that need to be fooled into where a file is located. Of course the real answer is to fix the application, but once again the conventions of the MS Windows environment are that only the software vendor can do that.

HDD reliability is not a constant anymore

[Anonymous Coward]

People are only talking about SSD reliability. Since the floods in Thailand I have found HDD reliability to be worse.

Oh no..it's RAID!

[doccus]

I would have thought that SSDs are ideal candidates for integration into a HHD/SSD hybrid RAID formation. The problem with how SSDs go, is there's no warning. One day, they just.. stop. We all know about the ""click of death" in a standard HHD, whch usually gives us enough time to run to the shop and find the only available drive has half the capacity of the one that's failing ;-). We do get time to get the data off , though... Not so with SSDs..If I ever get one I fiully intend to try this hybrid RAID idea.. the best configuration I haven't sorthed out yet, though...

Re:Oh no..it's RAID!

[doccus]

Also, if I coud choose particulars on where, and how much, memory is alocated to VM on OSX, I would definitely use an SSD for that.. I seem to recall seeing hybrid drives that do something like that aanyways, by loading the currently used data off the HD onto the SSD portion....

Re:Oh no..it's RAID!

[Guy Harris]

Also, if I coud choose particulars on where, and how much, memory is alocated to VM on OSX, I would definitely use an SSD for that...

Presumably by "memory allocated to VM" you mean "swap space".

You don't get to control how much - dynamic_pager will add swap files whenever the kernel tells it to, and the kernel will tell it to do so whenever there's not enough swap space, although the kernel will also tell it to delete swap files if there's enough unused swap space.

You can, however, specify the directory in which they're created, by using the -F flag to dynamic_pager. See the dynamic_pager man page and the /System/Library/LaunchDaemons/com.apple.dynamic_pager.plist property list. (Change it and you'll need to reboot.)

I'm not sure whether this would work if the drive in question isn't the drive with the root partition and diskarbitrationd doesn't mount the relevant partition before dynamic_pager starts. You might be able to say "launch this once this directory exists".

Re:Oh no..it's RAID!

[doccus]

Yes, you are correct, of course, swap space.. the allocated memory is up to the kernel.. Every other OS allows you to select a swap partition however, and it is this that an SSD would be ideal for...

Re:Oh no..it's RAID!

[Guy Harris]

Yes, you are correct, of course, swap space.. the allocated memory is up to the kernel.. Every other OS allows you to select a swap partition however

Presumably by "a swap partition" you mean "a partition that contains swap space in some fashion", as Windows (NT and possibly OT) also use a file as backing store and don't, as far as I know, allow a raw partition to be used in that fashion.

If that's what you mean, then, as per my post, OS X allows you to specify the directory in which to put swap files, and, if the directory can be on a volume other than the root volume, would let you swap to another drive.

Re:Oh no..it's RAID!

[doccus]

Well, I just habitually allow for the file to acheive maximum size as per the partition.. I don't like having a swap file on a partition that is EVER used for another purpose as well. I'll check out the tips you mentioned however. THX

Re:Oh no..it's RAID!

[Guy Harris]

Well, I just habitually allow for the file to acheive maximum size as per the partition.. I don't like having a swap file on a partition that is EVER used for another purpose as well.

I guess you could use the -S flag for that:

-S The fixed filesize [in bytes] to use for the paging files. By
default dynamic_pager uses variable sized paging files, using
larger sized files as paging demands increase. The -S, -H and -L
options disable that default and cause dynamic_pager to use a
series of fixed sized external paging files.

Re: and by size and warrantee?? (50-100%^^)

[lpq]

The summary also said they explored failures under warrantee. Have 80GB SSD's even been around as long as a 5-yr-warrantee HD?

How do the SSD's stack up in failures/GB?

Face it, at 98.5% failure rate and 6-8 drives = 1 HD, we are talking,
um... 98.5**6 - 98.5**8 = 91.3% - 88.6% chance of NO failure, or
8.7-11.4% chance of failing in a shorter warrantee period for the same
amount of disk space or 50-100% higher.