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.

Desktop Iometer - 4KB Random Read (4K Aligned)

Random read performance is staggering - a good 40% higher than anything else we've tested. While the cutoff for usefulness on a client drive is likely much lower than what even the Octane could deliver, this sort of performance bodes very well for OCZ's enterprise ambitions.

Desktop Iometer - 4KB Random Write (4K Aligned) - 8GB LBA Space

Randomly write performance is also just excellent. SandForce's peak numbers come close, but throw in any sort of incompressible data and they quickly take a step back while the Vertex 4 is able to deliver. Again, I'm actually more interested in these numbers from an enterprise workload standpoint but heavy client users will definitely not be disappointed.

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:

Desktop Iometer - 4KB Random Write (8GB LBA Space QD=32)

SandForce always scaled well at higher queue depths, but again we're looking at best case performance for the SF-2281. Move towards incompressible data and the Vertex 4 is alone at the top.

Inside the Vertex 4 Sequential Read/Write Speed
POST A COMMENT

126 Comments

View All Comments

  • jwilliams4200 - Saturday, April 07, 2012 - link

    There is no hard evidence that is publicly available since OCZ is keeping it quiet for obvious reasons.

    The Octane uses a rebadged Marvell 88SS9174, and the Vertex 4 uses a rebadged Marvell 88SS9187.
    Reply
  • rw1986 - Saturday, April 07, 2012 - link

    I'm just curious how you can make these claims so authoritatively when, as you say "there is no hard evidence that is publicly available" to support your position.

    What you seem to be suggesting is that OCZ made a 32 million dollar acquisition in April of 2011 (Indilinx) and then decided to simply license and rebadge controllers from Marvell rather than build internally (which is completely contrary to what OCZ has told investors and the public). On a Thursday conference call OCZ's CEO was quite adamant that Everest 2 is completely their IP and will improve their profit margin strucutre as a result (since they don't have to pay a fee to Sandforce for each controller as before). If what you are saying is true then OCZ has misled their investors at best.

    This is a pretty serious accusation you are making here and it would be nice if you had a little more to say in support of it than "trust me, i know." If you were really in a position to know then I doubt you would be posting on the comments section at AT.

    I'd be interested to hear some experts weigh in on this exchange. How could we verify or disprove that the Everest 2 is really not Indilinx but instead a rebadged 9187? Surely some analysis can be done on the ASIC
    Reply
  • jwilliams4200 - Saturday, April 07, 2012 - link

    It isn't very important to me either way, since I have no financial interest. But OCZ has a history of dishonesty, and it bothers me to see them get away it.

    If it is important to you, you could start by contacting the appropriate people at OCZ and asking whether the controllers in the Octane and the Vertex 4 have similar or identical circuits with Marvell controllers.
    Reply
  • jwilliams4200 - Wednesday, April 11, 2012 - link

    http://www.anandtech.com/show/5741/ocz-confirms-oc... Reply
  • hackztor - Friday, April 13, 2012 - link

    Good job on being correct. I think alot of people feel betrayed again by ocz. Last time i will purchase their product no matter what. Reply
  • pookguy88 - Wednesday, April 04, 2012 - link

    so wait, right now they'll (256gb, 512gb) ship with 1gb DRAM but later on 512mb? Is it just me or is that going to be a huge deal for customers once they realize that they basically got jacked 512mb of RAM vs early adopters? Reply
  • pattycake0147 - Wednesday, April 04, 2012 - link

    I inferred that the smaller drives will ship with 1GB but only 512MB will be used on the drive. Reply
  • Voo - Wednesday, April 04, 2012 - link

    Considering that was explicitly stated in the article there's not much to infer here is it?

    "Oh no I was jacked from 512mb that were deactivated anyhow - what a scandal!"
    Reply
  • antef - Wednesday, April 04, 2012 - link

    Why is your standard recommendation still the Samsung SSD 830? Given the performance, prices, and most notably, the incredibly excellent reviews on sites like Newegg, the Crucial M4 would still be my recommendation to new shoppers. Reply
  • Kristian Vättö - Wednesday, April 04, 2012 - link

    Samsung 830 is "a lot" faster, see our Heavy and Light benches. In real world, you will probably not notice the difference unless you do something I/O intensive. Personally, I've been recommending both Crucial m4 and Samsung 830. Crucial is great if you want to save a few bucks but otherwise Samsung is better. I think Anand has been using the Samsung in his main system for months now and haven't had issues, so that may be a reason as well. After all, we speak based on our own experience. Reply

Log in

Don't have an account? Sign up now