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

Here you can see the cap on 4KB random writes alive and well. As I've mentioned in previous articles, we're finally good enough when it comes to 4KB random write performance for current desktop workloads - so despite the cap you won't see any real world impact of it in our tests.

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

Iometer - 4KB Random Read, QD=3

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

Iometer - 128KB Sequential Read

Introduction AnandTech Storage Bench 2011
POST A COMMENT

45 Comments

View All Comments

  • Concillian - Thursday, May 05, 2011 - link

    OCZ is still SandForce's favorite partner and thus it gets preferential treatment when it comes to firmware.


    This is why I won't buy a SandForce SSD. Yeah, I can get a brand that doesn't have a cap. Or I can go with a different SSD that doesn't force me to jump through hoops to make sure I'm the same hardware from the right vendor.

    The same hardware from different vendors should not have vastly different performance. How many people would put up with a memory bandwidth limit on P67 chipset motherboards from Gigabyte, but not ASUS? (or whatever brands.) No memory bandwidth doesn't have a huge impact on overall PC performance, but I think it would still be a big deal if something like that actually happened.

    The SF-2281 either needs all vendors capped or none. It's a really shady tactic to offer two versions of the same hardware IMO.
    Reply
  • semo - Thursday, May 05, 2011 - link

    Nobody in the know likes the SF games but the whole thing is so complicated that most people won't understand it. Suits OCZ as the recent bad publicity doesn't seem to have affected them and everyone thinks they are the best choice for SSDs.

    Where are the Corsair force GT drives? Also, why are there no reviews of the Samsung 470?
    Reply
  • Mr Perfect - Thursday, May 05, 2011 - link

    I agree. If you want to offer vendors special products, fine, but give them a different model number. Call a controller capable of 27k a 2280 and the 52k version the 2281. You can still have incentive products, but the consumer doesn't get duped. Everyone's happy. Reply
  • Flunk - Thursday, May 05, 2011 - link

    Almost all IC vendors do this. Intel is probably the worst by far. Selling essentially the same chip up to 50 different ways but lasering off parts of it. Almost all onboard sound chips, network chips, drive controllers, GPUs and anything else you can think of uses the same strategy. Reply
  • Chloiber - Thursday, May 05, 2011 - link

    So the huge IOPS are pretty much useless, if the QD needs to be high - which is the case with every SSD.

    Anand, how is the "burst" rate of the Mercury regarding Random Write IOPS? I remember that with SF 12xx, the burst rate was exactly the same (for some seconds), only after 5-20s you could see a difference between the "unlocked" Vertex 2 and the rest. Considering how often one needs the random write performance for several seconds or even minutes (= never) I still think those huge IOPS numbers and the "unlocked firmware" stuff are just a huge marketing stunt. The benefit for the "normal" home user is = zero.
    Reply
  • semo - Thursday, May 05, 2011 - link

    Why is a high IOPS figure useless? Just because average Joe facebook doesn't do continuous IO intensive operations doesn't mean we don't need fast SSDs. You can apply your "logic" to CPU, GPU and pretty much any other technological advancement. Reply
  • kmmatney - Thursday, May 05, 2011 - link

    For the normal home user, the Anand light workload test is really the best thing to look at - no need to look at any other metric. The drive does really well here. Reply
  • Robear - Thursday, May 05, 2011 - link

    I believe most people who are interested in the power consumption are most interested in how it performs in a notebook. 2W versus 7W in power is negligible on a desktop. Instead of using a Velociraptor, can you please compare the SSD to a notebook hard drive, like maybe a Seagate Momentus? Reply
  • krazyderek - Thursday, May 05, 2011 - link

    the momentus XT is included, the XT was a little more power hungry then typical notebook drives, have a look at the past review for more info to compare

    http://www.anandtech.com/show/3734/seagates-moment...

    looks like some the new round of SSD's forgo power savings to move up the performance latter (ie: 240gb OCZ V3)
    Reply
  • mschira - Thursday, May 05, 2011 - link

    I was wondering if one could fit this drive into a 7mm slimline slot such as the Lenovo T420s.
    Lenovo only offers an Intel 160gb drive but I would fancy the possibility to insert a speedier 240gb SSD. maybe when removing some of the casing?
    cheers
    M.
    Reply

Log in

Don't have an account? Sign up now