Random Read/Write Speed

The four corners of SSD performance are as follows: random read, random write, sequential read and sequential write speed. Random accesses are generally small in size, while sequential accesses tend to be larger and thus we have the four Iometer tests we use in all of our reviews.

Our first test writes 4KB in a completely random pattern over an 8GB space of the drive to simulate the sort of random access that you'd see on an OS drive (even this is more stressful than a normal desktop user would see). I perform three concurrent IOs and run the test for 3 minutes. The results reported are in average MB/s over the entire time. We use both standard pseudo randomly generated data for each write as well as fully random data to show you both the maximum and minimum performance offered by SandForce based drives in these tests. The average performance of SF drives will likely be somewhere in between the two values for each drive you see in the graphs. For an understanding of why this matters, read our original SandForce article.

Iometer - 4KB Random Write, 8GB LBA Space, QD=3

Peak performance on the 120GB Vertex 3 is just as impressive as the 240GB pre-production sample as well as the m4 we just tested. Write incompressible data and you'll see the downside to having fewer active die, the 120GB drive now delivers 84% of the performance of the 240GB drive. In 3Gbps mode the 240 and 120GB drives are identical.

Many of you have asked for random write performance at higher queue depths. What I have below is our 4KB random write test performed at a queue depth of 32 instead of 3. While the vast majority of desktop usage models experience queue depths of 0 - 5, higher depths are possible in heavy I/O (and multi-user) workloads:

Iometer - 4KB Random Write, 8GB LBA Space, QD=32

At high queue depths the gap between the 120 and 240GB Vertex 3s grows a little bit when we're looking at incompressible data.

Iometer - 4KB Random Read, QD=3

Random read performance is what suffered the most with the transition from 240GB to 120GB. The 120GB Vertex 3 is slower than the 120GB Corsair Force F120 (SF-1200, similar to the Vertex 2) in our random read test. The Vertex 3 is actually about the same speed as the old Indilinx based Nova V128 here. I'm curious to see how this plays out in our real world tests.

Sequential Read/Write Speed

To measure sequential performance I ran a 1 minute long 128KB sequential test over the entire span of the drive at a queue depth of 1. The results reported are in average MB/s over the entire test length.

Iometer - 128KB Sequential Write

Highly compressible sequential write speed is identical to the 240GB drive, but use incompressible data and the picture changes dramatically. The 120GB has far fewer NAND die to write to in parallel and in this case manages 76% of the performance of the 240GB drive.

Iometer - 128KB Sequential Read

Sequential read speed is also lower than the 240GB drive. Compared to the SF-1200 drives there's still a big improvement as long as you've got a 6Gbps controller.

The Vertex 3 120GB AnandTech Storage Bench 2011
POST A COMMENT

154 Comments

View All Comments

  • spensar - Saturday, April 30, 2011 - link

    Love the real world benchmarks, and would like to see the Vertex 2 120gb numbers put up in the comparisions as well. Reply
  • jharmon - Monday, May 02, 2011 - link

    It's been almost a month since an update in the SSD world from Anandtech. Are there any reviews on the near horizon? You said one might want to wait for a purchase until you got some of the lower capacity drives in to test to compare with the vertex 3.

    Thanks for all your hard work in analysis, Anand!
    Reply
  • jharmon - Monday, May 02, 2011 - link

    Maybe also address the issue with the hardware limitation from the Marvell controllers 91xx. If I understand it correctly, these will be be able to achieve the 500 MB/s. If that performance cannot be achieved, is the vertex 3 worth the premium? Reply
  • mrkimrkonja - Wednesday, May 04, 2011 - link

    From these SSD articles I learned that TRIM is very important with SSD.
    I read somewhere that when you use SSD in RAID you lose TRIM support and some other things.
    Is this true?
    I now have 2 old ADATA SSD generation 1 in RAID0 they have 2x the performance comparing to a single drive bud they do not have TRIM anyway.
    I already bought one VERTEX 2 60GB and thought buying anther one.
    Always thought better more smaller disks in RAID than one big one.
    If this is true I an thinking I would lose more over time without TRIM support than gain with RAID0.

    Can you hal me with some info or first hand tests?
    Reply
  • Foochey - Wednesday, May 04, 2011 - link

    I would be very interested to see how the 480GB stacks up to these drives. Since these are parallel processing devices, you would think that the 480 would perform even better than the 240. Anand, any way you can get a 480 and test it? OCZ's specs look the same, so I'm guessing that they are using the smaller die chips or doing something with the way they write to the array. Any ideas? Price certainly makes the 480 out of the question for most of us, but I sure would be interested in its performance. Reply
  • jeffburg - Thursday, May 26, 2011 - link

    OCZ Just launched the Max IOPS Version. Is that worth the extra $10? Whats the difference between the two? Reply
  • Palen - Thursday, June 02, 2011 - link

    Thank you for another great article.

    In the conclusion of this article it's advised to wait for a couple of weeks to see how the vertex 3 120 GB goes againts the other 3rd generation 120GB SSD's. Is there any indication that Anandtech will be recieving / testing any of these drives in the near future?

    Another thing is RAID: I've been doing some digging on SSD's, since I didn't know much about them untill last week. The general picture is starting to make sence now. Bigger is ussually faster (within certain parameters). However i'm getting conflicting information about RAID. According to some it's worth it, while others urge to stay away from RAID (no TRIM support, etc). It would be nice to have a well respected opinion in the mass of conflicting information. How would for instance 2x 60 GB SSD's do against a same type/brand 120 GB SSD? and how does the raidcontroller fit into this equation? On-board vs. professional solution?

    My gut says "stay away from raid!" because the SSD's allready scale in performance depending on their storagesize. Next to that the strain that an onboard controller will put on your CPU. And paying for a professional raidcard just seems silly, with the range of SSD PCIe storage solutions. But than again, what do I know?
    Reply
  • paul-p - Saturday, October 22, 2011 - link

    After 6 months of waiting for OCZ and Sandforce to fix their firmware from freezes and BSOD's, I can finally say it is fixed. No more freezes, no more BSOD's, performance is what is expected. And just to make sure all of the other suggestions were 100% a waste of time, I updated the firmware and DID NOT DO anything else except reboot my machine and magically everything became stable. So, after all these months of OCZ and Sandforce blaming everything under the sun including:

    The CMOS battery, OROM's, Intel Drivers, Intel Chipsets, Windows, LPM, Hotswap, and god knows what else, it turns out that none of those issues had anything to do with the real problem, which was the firmware.

    While I'm happy that this bug is finally fixed, Sandforce and OCZ have irrepairably damaged their reputation for a lot of users on this forum.

    Here is a list of terrible business practices that OCZ and Sandforce have done over the last year...

    OCZ did not stand behind their product when it was clearly malfunctioning is horrible.
    OCZ did not allow refunds KNOWING that the product is defective is ridiculous.
    OCZ nor Sandforce even acknowledged that this was a problem and steadfastly maintained it only affected less than 1% of users.
    The fact that OCZ claims this bug affected 1% of users is ridiculous. We now know it affected 100% of the drives out there. Most users just aren't aware enough to know why their computer froze or blue screened.
    OCZ made their users beta test the firmwares to save money on their own testing
    OCZ did not have a solution but expected users to wipe drives, restore from backups, secure erase, and do a million other things in order to "tire out" the user into giving up.
    OCZ deletes and moves threads in order to do "damage control and pr spin".

    But the worst sin of all is the fact that it took almost a year to fix such a MAJOR bug.

    I really hope that OCZ learns from this experience, because I'm certain that users will be wary of Sandforce and OCZ for some time. It's a shame, because now that the drive works, I actually like it.
    Reply
  • paul-p - Saturday, October 22, 2011 - link

    I just want to thank Anandtech for being the ONLY site out there there called out OCZ and Sandforce for having defective products. While every other hardware site out there were kissing OCZ and Sandforces butt and saying that the Vertex 3 SSD was the best thing since sliced bread, Anandtech actually was one of the only sites that actually acknowledged that there was a major bug with the sandforce controllers.

    I can't believe all the review sites out there were praising the Vertex 3 for almost a year when the drive had major BSOD and freezing issues. A review should be more than running a benchmark on the drive, but checking to see how the drive performs and making sure it is stable. In my eyes, every other review site showed me that they care more about the sponsors and advertising dollars than the users that visit their site. So, once again, thanks Anandtech for speaking the truth when others wouldn't.
    Reply
  • danwat12345 - Sunday, November 13, 2011 - link

    I think I'll keep my 80GB Intel X25-m G2 SSD. From your benchmarks it looks like the 120GB Vertex 3 over SATA 2 isn't that much better with random read operations than my trusty ole' 80GB Intel.

    I am curious if your Anandtech storagebench 2011 test were done on completely in-compressible data? The random read numbers of the Vertex 3 120GB SATA2 drive wasn't very impressive.
    Reply

Log in

Don't have an account? Sign up now