The SSD Relapse: Understanding and Choosing the Best SSD
by Anand Lal Shimpi on August 30, 2009 12:00 AM EST- Posted in
- Storage
One Tough Act to Follow
What have I gotten myself into? The SSD Anthology I wrote back in March was read over 2 million times. Microsoft linked it, Wikipedia linked it, my esteemed colleagues in the press linked it, Linus freakin Torvalds linked it.
The Anthology took me six months to piece together; I wrote and re-wrote parts of that article more times than I'd care to admit. And today I'm charged with the task of producing its successor. I can't do it.
The article that started all of this was the Intel X25-M review. Intel gave me gold with that drive; the article wrote itself, the X25-M was awesome, everything else in the market was crap.
Intel's X25-M SSDs: The drives that started a revolution
The Anthology all began with a spark: the SSD performance degradation issue. It took a while to put together, but the concept and the article were handed to me on a silver platter: just use an SSD for a while and you’ll spot the issue. I just had to do the testing and writing.
OCZ's Vertex: The first Indilinx drive I reviewed, the drive that gave us hope there might be another.
But today, as I write this, the words just aren't coming to me. The material is all there, but it just seems so mature and at the same time, so clouded and so done. We've found the undiscovered country, we've left no stone unturned, everyone knows how these things work - now SSD reviews join the rest as a bunch of graphs and analysis, hopefully with witty commentary in between.
It's a daunting, no, deflating task to write what I view as the third part in this trilogy of articles. JMicron is all but gone from the market for now, Indilinx came and improved (a lot) and TRIM is nearly upon us. Plus, we all know how trilogies turn out. Here's hoping that this one doesn't have Ewoks in it.
What Goes Around, Comes Around
No we're not going back to the stuttering crap that shipped for months before Intel released their X25-M last year, but we are going back in the way we have to look at SSD performance.
In my X25-M review the focus was on why the mainstream drives at the time stuttered and why the X25-M didn't. Performance degradation over time didn't matter because all of the SSDs on the market were slow out of the box; and as I later showed, the pre-Intel MLC SSDs didn’t perform worse over time, they sucked all of the time.
Samsung and Indilinx emerged with high performance, non-stuttering alternatives, and then we once again had to thin the herd. Simply not stuttering wasn't enough, a good SSD had to maintain a reasonable amount of performance over the life of the drive.
The falling performance was actually a side effect of the way NAND flash works. You write in pages (4KB) but you can only erase in blocks (128 pages or 512KB); thus SSDs don't erase data when you delete it, only when they run out of space to write internally. When that time comes, you run into a nasty situation called the read-modify-write. Here, even to just write 4KB, the controller must read an entire block (512KB), update the single page, and write the entire block back out. Instead of writing 4KB, the controller has to actually write 512KB - a much slower operation.
I simulated this worst case scenario performance by writing to every single page on the SSDs I tested before running any tests. The performance degradation ranged from negligible to significant:
| PCMark Vantage HDD Score | New | "Used" |
| Corsair P256 (Samsung MLC) | 26607 | 18786 |
| OCZ Vertex Turbo (Indilinx MLC) | 26157 | 25035 |
So that's how I approached today's article. Filling the latest generations of Indilinx, Intel and Samsung drives before testing them. But, my friends, things have changed.
The table below shows the performance of the same drives showcased above, but after running the TRIM instruction (or a close equivalent) against their contents:
| PCMark Vantage HDD Score | New | "Used" | After TRIM/Idle GC | % of New Perf |
| Corsair P256 (Samsung MLC) | 26607 | 18786 | 24317 | 91% |
| OCZ Vertex Turbo (Indilinx MLC) | 26157 | 25035 | 26038 | 99.5% |
Oh boy. I need a new way to test.
Facebook
Twitter
RSS
295 Comments
View All Comments
Anand Lal Shimpi - Monday, August 31, 2009 - link
wow I misspelled my own name :) Time to sleep for real this time :)Take care,
Anand
IntelUser2000 - Monday, August 31, 2009 - link
Looking at pure max TDP and idle power numbers and concluding the power consumption based on those figures are wrong.Look here: http://www.anandtech.com/cpuchipsets...px?i=3403&a...">http://www.anandtech.com/cpuchipsets...px?i=3403&a...
Modern drives quickly reach idle even between times where the user don't even know and at "load". Faster drives will reach lower average power because it'll work faster to get to idle. This is why initial battery life tests showed X25-M with much higher active/idle power figures got better battery life than Samsungs with less active/idle power.
Max power is important, but unless you are running that app 24/7 its not real at all, especially the max power benchmarks are designed to reach close to TDP as possible.
Anand Lal Shimpi - Monday, August 31, 2009 - link
I agree, it's more than just max power consumption. I tried to point that out with the last paragraph on the page:"As I alluded to before, the much higher performance of these drives than a traditional hard drive means that they spend much more time at an idle power state. The Seagate Momentus 5400.6 has roughly the same power characteristics of these two drives, but they outperform the Seagate by a factor of at least 16x. In other words, a good SSD delivers an order of magnitude better performance per watt than even a very efficient hard drive."
I didn't have time to run through some notebook tests to look at impact on battery life but it's something I plan to do in the future.
Take care,
Anand
IntelUser2000 - Monday, August 31, 2009 - link
Thanks, people pay too much attention to just the max TDP and idle power alone. Properly done, no real apps should ever reach max TDP for 100% of the duration its running at.cristis - Monday, August 31, 2009 - link
page 6: "So we’re at approximately 36 days before I exhaust one out of my ~10,000 write cycles. Multiply that out and it would take 36,000 days" --- wait, isn't that 360,000 days = 986 years?Anand Lal Shimpi - Monday, August 31, 2009 - link
woops, you're right :) Either way your flash will give out in about 10 years and perfectly wear leveled drives with no write amplification aren't possible regardless.Take care,
Anand
cdillon - Monday, August 31, 2009 - link
I gather that you're saying it'll give out after 10 years because a flash cell will lose its stored charge after about 10 years, not because the write-life will be surpassed after 10 years, which doesn't seem to be the case. The 10-year charge life doesn't mean they become useless after 10 years, just that you need to refresh the data before the charge is lost. This makes flash less useful for data archival purposes, but for regular use, who doesn't re-format their system (and thus re-write 100% of the data) at least once every 10 years? :-)Zheos - Monday, August 31, 2009 - link
"This makes flash less useful for data archival purposes, but for regular use, who doesn't re-format their system (and thus re-write 100% of the data) at least once every 10 years? :-)"I would like an input on that too, cuz thats a bit confusing.
GourdFreeMan - Tuesday, September 1, 2009 - link
Thermal energy (i.e. heat) allows the electrons trapped in the floating gate to overcome the potential well and escape, causing zeros (represented by a larger concentration of electrons in the floating gate) to eventually become ones (represented by a smaller concentration of electrons in the floating gate). Most SLC flash is rated at about 10 years of data retention at either 20C (68F) or 25C (77F). What Anand doesn't mention is that as a rule of thumb for every 9 degrees C (~16F) that the temperature is raised above that point, data retention lifespan is halved. (This rule of thumb only holds for human habitable temperatures... the exact relation is governed by the Arrhenius equation.)Wear leveling and error correction codes can be employed to mitigate this problem, which only gets worse as you try to store more bits per cell or use a smaller lithography process without changing materials or design.
Zheos - Tuesday, September 1, 2009 - link
Thank you GourdFreeMan for the additional input,But, if we format like every year or so , doesnt the countdown on data retention restart from 0 ? or after ~10 year (seems too be less if like you said temperature affect it) the SSD will not only fail at times but become unusable ? Or if we come to that point a format/reinstall would resolve the problem ?
I dont care about losing data stored after 10 years, what i do care is if the drive become ASSURELY unsusable after 10 year maximum. For drives that comes at a premium price, i don't like this if its the case.