Slashdot Mirror
Endurance Experiment Kills Six SSDs Over 18 Months, 2.4 Petabytes
crookedvulture writes Slashdot has previously covered The Tech Report's SSD Endurance Experiment, and the final chapter in that series has now been published. The site spent the last 18 months writing data to six consumer-grade SSDs to see how much it would take to burn their flash. All the drives absorbed hundreds of terabytes without issue, far exceeding the needs of typical PC users. The first one failed after 700TB, while the last survived an astounding 2.4 petabytes. Performance was reasonably consistent throughout the experiment, but failure behavior wasn't. Four of the six provided warning messages before their eventual deaths, but two expired unexpectedly. A couple also suffered uncorrectable errors that could compromise data integrity. They all ended up in a bricked, lifeless state. While the sample size isn't large enough to draw definitive conclusions about specific makes or models, the results suggest the NAND in modern SSDs has more than enough endurance for consumers. They also demonstrate that very ordinary drives can be capable of writing mind-boggling amounts of data.
Within a year, but not from the memory, but a shit controller. From OCZ iirc. Something about it getting past half full giving it bad performance and anything before a full, drawn out format didn't cut it.
Since then, no SSDs died. But a fair number of spinning disks.
I think all most people are waiting for is for the GB/$ gap to decrease markedly. Otherwise it stays a SSD for your boot drive, a spinning disk to archive your junk market.
The fact that 2 of them died without warning is disappointing. I would rather have a shorter life time, but a clear indication that the drive is going to die.
Being an AC, I would chalk this up to a joke or trolling. But.... on the off chance that you are serious, I will bite.
Yes, you COULD use an SSD as swap, but it will not help THAT much. An SSD is much faster than a mechanical disk, but still a couple of orders of magnitude slower than real RAM. That upgrade would be like the difference between jogging with 50 pounds on your back, and then lowering it to 35 pounds. Yes, it will make a difference and make things better, but how much better to have no weight at all?
Just get more RAM. If your system cannot hold more RAM, then get a new mobo. If you regularly go over 16 GB of actual RAM in use, even going to a slower processor will be an improvement if you stop swapping. Hitting the swap file is a great way to make a fast processor do nothing for a while.
"-1 Troll" is the apparently the same as "-1 I disagree with you."
Talk about your planned obsolescence - not a single sector reallocation registered, but the firmware counter says it's write-tolerance is reached so it kills itself. I suppose it's nice that it switches to read-only mode when it dies, except for the fact that it bricks itself entirely after a power cycle. I mean come on - if it's my OS and/or paging drive then switching to read-only mode is going to kill the OS almost immediately, and there goes my one chance at data recovery. Why not just leave it in permanent read-only mode instead? Sure it's useless for most applications, but at least I can recover my data at my leisure.
--- Most topics have many sides worth arguing, allow me to take one opposite you.
Who thought this was a good idea? If the drive thinks future writes are unstable, good for it to go into read only mode. But to then commit suicide on the next reboot? What if I want to take one final backup, and I lose power?
As SSD cells wear, the problem is that they hold charge for less time. Starting new, the time that the charge will be held would be years, but as the SSD wears, the endurance of the held charge declines.
Consequently, continuous write tests will continue to report "all good" with a drive that is useless in practice, because while the continuous write will re-write a particular cell once every few hours, it might only hold a charge for a few days - meaning if you turned it off for even a day or so, you'd suffer serious data loss.
SSDs are amazing but you definitely can't carry conventional wisdom from HDDs over.
I have no problem with your religion until you decide it's reason to deprive others of the truth.
the results suggest the NAND in modern SSDs has more than enough endurance for consumers
"Challenge accepted." - some guy trying to invent octo-level-cell flash
Wanna know how to kill a spinning disk? Put it in a DVR. My DVR (with the "pause and rewind live TV" ability), would re-write 100% of the time. It died in a few months. Replaced it with a larger one and turned off the live-TV buffer, and it's lasted years. But it's all anecdotal, so I expect tests like this to give us some level of comparison.
Learn to love Alaska
Computers with memory leaks?
John McAfee 'It was like that time I hired that Bangkok prostitute; to do my taxes, while I fucked my accountant'
I'd like a mix of drives on my next box. A moderate "traditional" spinning oxide 1TB drive with a lot of cache for the primary boot, swap, and home directories, and an SSD mounted as my project workspace under my home directory. The work directory is where I do 99% of my writes, producing roughly 3GB for one particular project in about an hour's time.
My existing drive on my main box has survived a god-awful number of writes, despite it's 1TB size. My work is emphatically I/O bound for the past month or so, since I did some bug fixing and tuning, so switching that project directory over to an SSD would speed things up even for this aging 3.8 GHz P4 single core. But not enough to justify the investment in just an SSD without a boost in memory, CPU, and memory bandwidth.
I've got my eye on a Lenovo i7 unit for about $900, plus SSD, but it'll take about a year to save up the money. On the bright side, either the specs on the unit will improve or the price will come down with an intervening year before purchase. :D
I do not fail; I succeed at finding out what does not work.
I'm actually not convinced it's a terrible idea. It seems to me that when swapping really gets slow, its because of thrashing the hard disk. SSD may be several orders of magnitude slower than memory, but it's also several orders of magnitude faster than a thrashing HDD. It seems like it might be a reasonable tradeoff to make. New memory might require a new motherboard (and maybe a new processor), costing several hundred dollars, a not-insignificant amount of effort to swap it out, possibilities of new bugs with hardware incompatabilities, etc...all for some additional memory with, 99% of the, probably serves no additional benefit. On the other hand, it takes no more than a few minutes to move your swap between drives, costs nothing, and has almost no risk of any incompatibilities (unless you start ending up with applications fighting for throughput under the additional load).
Okay, those numbers that you quoted are very arbitrary, I'd like to see anything to back that up. The near-instantaneous seek time of an SSD compared to a mechanical disk ought to be a major factor when it comes to swap performance, far more so than throughput. In any case, there are many SSD-only systems now, in which case the swap space is on the SDD whether you like it or not, so there's certainly not an unreasonable thing to try.
Oh no... it's the future.
If all programs and operating systems were perfect, increasing RAM would help more than faster swap. But things aren't perfect and operating systems like to talk to swap 'just because'.
On my aging Mac Pro with 32 GB RAM I put a 60 GB SSS (left over from a laptop upgrade) in as swap. Seems to make a modest difference with Premiere, After Effects and Vue, especially renders. But it really isn't all that noticeable. Not as noticeable as increasing the RAM. But if you have an extra drive caddy and an extra SSD it's easy to try.
Faster! Faster! Faster would be better!
Sorry, market rules...
“He’s not deformed, he’s just drunk!”
Why Intel, why? We can all discuss whether the device should prematurely fail by some arbitrary software limit, but why BRICK it, as it can cause complete data loss!?
Instead, just set the drive to always boot in read-only mode, with secure erase being the only other allowed command. Then someone can recover their data and wipe the drive for good.
Intel doesn't have confidence in the drive at that point, so the 335 Series is designed to shift into read-only mode and then to brick itself when the power is cycled. Despite suffering just one reallocated sector, our sample dutifully followed the script. Data was accessible until a reboot prompted the drive to swallow its virtual cyanide pill.
-=Lothsahn=-
If you can't get more RAM (especially with the trend in newer laptops being to have soldered in chips), buy as large a SSD as possible that you can dedicate to swap. The reason is that this gives the drive more cells to wear-level the swapfile writes over, prolonging the drive's life.
There is an interesting sweet spot where you can get away with having just 2GB of RAM for normal office and web usage if you swap to SSD. A system with 4GB of RAM and a conventional hard disk feels much more sluggish, even though it rarely swaps. If you have an old system where you can't easily upgrade the RAM because it's expensive DDR2, try adding a solid state disk.
Not all swap usage is created equal: If your working set fits into RAM, you're not going to be slowed down much by using swap space.
As swap, it is nowhere near good as RAM, but it has one advantage -- SSD excels at random writes, which is what swap is usually doing, so just because of this, it is better than regular disk. To boot, if one has the bay for it in a desktop, it might just be worth tossing in a 100-200 gig drive and using it for swap, as well as possibly moving the OS's partition to it as well, although it is good to have a lot of free pages on a SSD to wear-level a swapfile.
Also, an SSD is several orders of magnitude cheaper than RAM, and makes regular disk access much faster. So for cost-benefit analysis, upgrade to SSD before trying to get a huge amount of RAM.
I'm a good cook. I'm a fantastic eater. - Steven Brust
SATA revision 3.0 = 6 Gbit/s
DDR3 - 1600 = 12800 MB/s
"MB" = Mega-BYTES, so multiply by 8 for bits/seconds
DDR3 - 1600 = 102400 Mbits/s
DDR3 - 1600 = 102.400 Gbits/s
So, the peak bandwidth is about 17 times faster!
Now, let's look at latency.
Typical DDR RAM latency is around 10 ns (give or take, but that is an average number)
Typical SSD latency is around 0.1 us, which is around 100 ns. About ten times more.
One more thing here about these numbers.... An SSD is **NOT** RAM. If you page, you have to get the data FROM the SSD and put it INTO your RAM. From there, the RAM must be read again. So, even IF your SSD were exactly the same speed as your RAM, it will still be slower because it must be copied into RAM first before it can be used.
As to whether it is unreasonable, that depends. It will not cost much to try, but still a rather bad idea if you do a LOT of swapping.
"-1 Troll" is the apparently the same as "-1 I disagree with you."
This experiment only documents the survivability of the NAND Flash itself, really. I've had two consumer SSDs and at least one SD fail completely for other reasons; they became completely un-usable, not just un-writable. In the case of the SSDs at least, I was told it was due to internal controller failure, meaning the NAND itself was fine but the circuits to control and access it were trashed. I suppose a platter-drive analog to that would be having the platters in mint condition with all data intact but the servo coil melted, or something.
Since I've only owned three consumer SSDs and two of those died from a mode of failure that wasn't even addressed by this experiment, what am I to make of the real value of the results? They certainly have no meaning for me, but YMMV.
Dual channel ram is like 25GB/s and a few ns of latency.
SSD is typically 0.5GB/s and 100,000 ns of latency.
I have been using a Samsung 840 (not pro) 120GB SSD as disk drive with a 3gb swap partition since May 2013. It is my work computer so it sees a lot of action (since my work laptop only has 4gb of ram it ends up using a lot of swap), the SSD did not fail on me yet. I can corroborate what some other people are saying here, it still gets dog slow when I need to use the swap, but I did not compare with a disk-based swap.
You got your SSD latency off by a factor of 1,000
Hmm, wikipedia suggests that a typical SSD access time is actually about 10us, so about 1,000x slower than DDR RAM, not 10x
However, Typical HDD acces times are around 10ms, or about 1,000x slower still.
So while an SSD page file will be much slower than RAM, it will also be much faster than a HDD. Moreover, page file access patterns typically involve megabyte-sized bock writes and kilobyte sized (single page) random reads - which play directly to the strengths of a SSD.
Certainly more RAM will potentially grant a much larger performance boost to RAM-starved applications, but an SSD is far cheaper and should, in theory at least, offer a dramatic improvement over a HDD for page files. Plus it offers many other performance boosts as well, which may make it a much more attractive upgrade option for someone on a budget.
--- Most topics have many sides worth arguing, allow me to take one opposite you.
Funny you should mention that:
http://arstechnica.com/gadgets/2013/11/once-great-ssd-manufacturer-ocz-filing-for-bankruptcy/
Who thought this was a good idea?
Probably the same person that had killbots shut down when they reach their preset kill limit. :)
Suddenly a bunch of SSD drives of all types and manufacturers have shown up on ebay. Coincidence? I think not.
If I trust the suicide, I suppose the upside would be that the drive can be safely tossed without worrying about the data on it.
With the proper equipment, I'm sure the data can be recovered. Still best to thoroughly destroy the drive.
Don't try to out wierd me, three-eyes. I get stranger things than you, free with my breakfast cereal. --Zaphod Beeblebr
You haven't actually tried it, have you? Putting a swap file on a SSD instead of HDD helps tremendously.
It's not because the SSD transfers data at a faster speed. It's because its seek times are minuscule. A typical HDD's seek time is around 10 ms. A SSD's is around 0.1 ms, making the SSD 100x faster. In practical terms, this means when the computer starts swapping to a HDD, you cannot use the HDD for anything else. Were you opening a file when it started swapping? The computer's not gonna complete the read until it finishes swapping. Does the computer need to get one itsy bitsy file off the HDD to continue with it's current operation? It's gonna have to wait for the swap to finish. The swap writing to the HDD saturates the drive's seek capability and the drive cannot do anything else.
With a SSD, the "seek time" (actually time to find and read/write electronically) is so short the limiting factor is actually filesystem overhead. That's why enabling NCQ speeds up 4k read/writes on SSDs by about 10x, while it gives almost no speedup to HDDs. In practical terms, this means the computer can continue to use the SSD while it's swapping. My previous laptop came with 4GB of RAM and a SSD. I found I needed 8GB when using Photoshop, but the behavior while swapping was so amicable that I put off buying the RAM upgrade for 3 months until there was a good sale. I could tell it was swapping because it got a bit slower, but it wasn't show-stopping like with a HDD.
Even with HDDs, the recommendation if you had multiple drives was to put the OS and swap file on different physical drives for this reason. If you've ever used a system set up like that, swapping was nowhere near as bad as with a single HDD.
Some additional info from an earlier article:
According to Intel, this end-of-life behavior generally matches what's supposed to happen. The write errors suggest the 335 Series had entered read-only mode. When the power is cycled in this state, a sort of self-destruct mechanism is triggered, rendering the drive unresponsive. Intel really doesn't want its client SSDs to be used after the flash has exceeded its lifetime spec. The firm's enterprise drives are designed to remain in logical disable mode after the MWI bottoms out, regardless of whether the power is cycled. Those server-focused SSDs will still brick themselves if data integrity can't be verified, though.
SMART functionality is supposed to persist in logical disable mode, so it's unclear what happened to our test subject there. Intel says attempting writes in the read-only state could cause problems, so the fact that Anvil kept trying to push data onto the drive may have been a factor.
All things considered, the 335 Series died in a reasonably graceful, predictable manner. SMART warnings popped up long before write errors occurred, providing plenty of time—and additional write headroom—for users to prepare.
So, it sounds like this is the intended behavior for *enterprise* drives. It may not be the same for *consumer* drives, but that's a bit unclear.
While it may make you feel better if consumer SSD drives would go into a permanent read-only mode, it seems extremely unlikely that a typical consumer would ever actually reach this point in an SSD's life at all. So, I'm not really losing sleep that my own Intel SSD drives are going to brick themselves, when at a typical consumer write volume, this isn't going to happen anytime in the next century (seriously, look at the volume of data that was written). The drive will long be dead because of some electronic component failure long before I reach it's natural end of write life. Moreover, I'd appear to have plenty of warnings and could easily replace them long before that happened.
Irony: Agile development has too much intertia to be abandoned now.
Yes, swapping is better on a SSD. But, it is MUCH MUCH better to not swap at all. That is my point. If you have to have swap, you are better off just buying more RAM.
"-1 Troll" is the apparently the same as "-1 I disagree with you."
. In any case, there are many SSD-only systems now, in which case the swap space is on the SDD whether you like it or not, so there's certainly not an unreasonable thing to try.
The software that comes with my Samsung disables the windows swapfile if you want it to. Since I have plenty of RAM, that's okay with me.
Therefore, by the (faulty) logic you're using, you're just a cow with a keyboard - osu-neko (2604)
they don't spin much at all
Yes. I've done exactly this, on both cheap ultrabooks with 4GB ram and huge Linux servers with 512GB of RAM. (We have a 2.5TB Redis cluster that was running out of space waiting for additional nodes to be commissioned.)
It works. It works well. It's not a panacea, but it's an enormous improvement over swapping to spinning disk. Night and day.
No, really, I've done it. Loaded up a 4GB ultrabook with lots of applications and Chrome tabs and a couple of Virtualbox CentOS instances, over 6GB in active use. Switching apps initially took a couple of seconds as it settled down to a realistic working set, but after than that you couldn't tell that it was swapping at all.
I've done it on spinning disk too, of course, and that couple of seconds was closer to a minute.
As long as you're not actively thrashing - as long as your working set still fits in RAM - swapping to SSD is pretty much painless.
Buy a Corsair's Neutron SSD
Slashdot, fix the reply notifications... You won't get away with it...
That was the whole idea behind Microsoft's Readyboost, which allowed you to plug in a USB (!) thumbdrive, and if Windows deemed it fast enough, it would use it as swap space rather than (or in addition to) the hard drive. My experience is that actually worked somewhat well, given a laptop with a 5400RPM drive and a hardware limitation of 3GB of ram. Though I eventually ditched it when I replaced the HDD with a SSD, which made a tremendous difference.
Interestingly, Readyboost won't let you use a SSD hooked up over SATA (I tried just to see if it would). Though it will let you use a card reader so long as it deems the card fast enough.
This reminds me of the systems they tried to impose in the early 70s' US cars: if the seat belts weren't latched, the cars would not start. No, not just an annoying bong for a brief time; the freaking engine would not even turn over. And this being completely analog, system failures happened quite a bit. But fear not! If there was a failure, there was an under hood button you could push to bypass the system. Once. A one-time use, then you had to tow the car to a dealership for a new box. As most everyone by-passed this insanity, the system was quickly dropped.
Here's hoping Intel also comes to their senses about their one-time recovery window (hopefully with some well-deserved threats from the FTC).
If your page file is large enough, probably such that your total virtual memory is not over 80% utilization, the OS does not need to write data every time is pages out. A large number of pages rarely change. The OS only need to write out dirty pages, page that have changed. If the OS needs to make room and a page is not dirty, it can just remove it from memory and read it from disk later, no writes needed.
Making your page file large enough could reduce writes or just get enough memory. I do find it interesting that just loading video games can cause a considerable amount of writes. World of Warcraft writes nearly 1MB/s while reading 500MB/s. Over the period of 10 seconds, that's 10MB of data written. What could WoW be writing 10MB of? Maybe it's NTFS. I should check if NTFS is updating last time accessed.
SSDs may be 10us, but over SATA3.0, your real world is more like 100us.
You can never fully disable the Windows swap, it is hard coded into the kernel that is must exist or the kernel will not work. Even if it claims to be disabled, it still creates a 32MB page file on the boot drive, and can thrash the heck out of it. Win8.1 won't even let me reduce the page file below 800MB. I tried disabling it and it said it'll still create an 800MB page file.
You also have issues with memory fragmentation. If there is not a large enough contiguous free memory address, the OS can page out and do some minor defragging. It is relatively simple to contrive a situation where 50% of your memory is in use, but you'll get an out of memory error trying to allocate a page larger than 4KB. In the real world, the 80/20 rule is at play. Once you get past 80% usage, you could start running into memory issues with out a swap scratch pad.
I know this is slightly off-topic, but I found this surprising. About twenty years ago my main PC was a 66 MHz machine with 8 MB of RAM running Windows 95. I was learning to use the 3D graphics program "trueSpace" and I created a scene that was 11 MB big when saved to the hard drive as a wireframe. When I tried to render the scene, the hard drive thrashed for ten hours straight, and the scene was still only halfway rendered. Later, I bought 16 MB of RAM (if I recall for ~$400[!]), bringing my total up to 24 MB. That same scene rendered completely in twenty minutes. That was an fascinating lesson.
I've had some SSDs last for almost three years, but I would not trust them for important data. They are fine as a cache for speeding up OS access, or for a music player, but a magnetic hard drive is better for professional use.
You don't understand professional use. A professional would never, ever, ever trust a single device/system for important data. Not ZFS, not tape, not hard drives, not SSDs, not stone tablets.
SWM seeks new sig for a brief fling