The SSD Anthology: Understanding SSDs and New Drives from OCZ
by Anand Lal Shimpi on March 18, 2009 12:00 AM EST- Posted in
- Storage
Simulating a Used Drive
Since SSD performance degrades over time, it’s important to not only look at how well these drives perform new - but also the worst they’d perform over their lifetime. In order to do so we’d need a repeatable way of “seasoning” a drive to reduce its performance to the worst it could possibly get. The most realistic worst-case scenario is one where every single block on the drive is full of data. If a secure erase wipes all LBAs, that’s the best place to start. To simulate a well seasoned drive I first secure erased the drive.
After the secure erase, I used iometer to write one contiguous file across the disk - filling up the entire drive with 128KB blocks. In the case of the 80GB Intel X25-M, that’s 74.5GB of data on the drive before I run a single benchmark. The spare area is left untouched.
Next, I take my test image and I restore it onto the partition with a sector by sector copy. The sequential file write made sure that data is stored in every page of the SSD, the test image restore adds a twist of randomness (and realism) to the data.
There are other ways to produce a drive in its well-used state, but this ends up being the most consistent and repeatable. To confirm that my little simulation does indeed produce a realistically worn drive I ran PCMark on three different drives: 1) a freshly secure-erased Intel X25-M, 2) an Intel X25-M setup using the method I just described and 3) the Intel X25-M used in my CPU testbed that has been through hundreds of SYSMark runs.
The benchmark of choice is PCMark Vantage; it simulates the real world better than most drive benchmarks. The results are below:
| Intel X25-M State | PCMark Vantage Overall Test | PCMark Vantage HDD Test |
| Fresh Secure Erase | 11902 | 29879 |
| Simulated Used Drive | 11536 | 23252 |
| Actual Testbed Used Drive | 11140 | 23438 |
The secure erased system loses about 3% of its overall performance and 22% of its hard drive specific performance compared to my testbed drive. The seasoning method I described above produces a drive with nearly identical drops in performance.
The method appears to be sound.
Now that we have a way of simulating a used drive, let’s see how the contestants fared.
Facebook
Twitter
RSS
250 Comments
View All Comments
blackburried - Wednesday, March 25, 2009 - link
It's referred to as "discard" in the kernel functions.It works very well w/ SSD's that support TRIM, like fusion-io's drives.
Iger - Wednesday, March 25, 2009 - link
This is the best review I've read in a very long time.Thank you very much!
BailoutBenny - Tuesday, March 24, 2009 - link
Great in depth article on flash based SSDs. I'm waiting for PRAM though.orclordrh - Tuesday, March 24, 2009 - link
Very illuminating article, very well written and researched. It made me glad that I didn't pull the trigger on an SSD for my I7 machine and regret not buying OCZ memory! I'm interested in adding an SSD as the scratch disk for Photoshop CS4 to use. I don't really launch applications very often, say once a week on the weekly reboot and keep 6-8 apps open at all times. I have 12GB of memory for that. The benchmarks were very interesting, but what sort of activity does Photoshop scratch usage create? Large files or random writes? What type of SSD would be most cost effective here?An SSD does sound better than a SSD!
semo - Wednesday, March 25, 2009 - link
wait for ddr3 to enter the mainstream and buy loads of memory.use a ramdisk for your adobe scratch area. much faster than ssd and no wear to worry about (not that you would worry that much with modern ssds anyway).
http://www.ghacks.net/2007/12/14/use-a-ramdisk-to-...">http://www.ghacks.net/2007/12/14/use-a-ramdisk-to-...
there is also a paid for and more feature rich ramdisk out there. can't remember the name
strikeback03 - Wednesday, March 25, 2009 - link
I'll have to check when I get home, but I believe the recommended size for the scratch disk is upwards of 10GB. So would need a motherboard that supports a LOT of RAM to give enough to main memory plus a scratch disk.strikeback03 - Wednesday, March 25, 2009 - link
I was wondering the same thing. I'd guess it would be a lot of writing/erasing, so an SSD might not be the best from a longevity standpoint, but if your system is hitting the scratch disk often then the speed might make it worthwhile.mikepers - Tuesday, March 24, 2009 - link
Anand,I wanted to compliment you on what I think was an excellent article. This is the type of thing I really have always liked from Anandtech. The detailed background, the technical reasons for the issues and then a thorough review of the current state of things.
I just finished upgrading my desktop. The only remaining item I wanted to replace was the hard disk. I had been thinking about getting a Velociraptor but instead I just ordered a 60GB Vertex from Newegg.
Thanks again for all the work.
Mike P.
ameatypie - Monday, March 23, 2009 - link
That sure was a lot to take in! Fantastic article though, it has really opened my eyes to the possibilities that Solid State Drives provide. Probably wont be buying one in the immediate future given the so-called depression and such things, but i will certainly keep up with SSD progress.Thanks again for your fantastic articles - im sure im not the only one who really appreciates them :)
coopchennick - Monday, March 23, 2009 - link
Hey Anand, I just finished reading through this whole article and I'm very impressed with the thoroughness and how informative it was.You just acquired a new regular reader.