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)

As we saw in our RevoDrive 3 X2 review, low queue depth random read performance doesn't really show much of an advantage on these multi-controller PCIe RAID SSDs. The Z-Drive R4 comes in a little faster than the RevoDrive 3 X2 but not by much at all. Even a single Vertex 3 does just fine here.

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

Random write performance tells a similar story, at such low queue depths most of the controllers aren't doing any work at all. Let's see what happens when we start ramping up queue depth however:

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

Surprisingly enough, even at a queue depth of 32 the Z-Drive R4 is no faster than the RevoDrive 3 X2. In fact, it's a bit slower (presumably due to the extra overhead of having to split the workload between 8 controllers vs just 4). In our RevoDrive review we ran a third random write test with two QD=32 threads in parallel, it's here that we can start to see a difference between these drives:

It's only at ultra high queue depths that the Z-Drive can begin to distance itself from the RevoDrive 3 X2. It looks like we may need some really stressful tests to tax this thing. The chart below represents the same data as above but in IOPS instead of MB/s:

271K IOPS...not bad.

The Card Sequential Read/Write Speed
POST A COMMENT

57 Comments

View All Comments

  • josephjpeters - Tuesday, September 27, 2011 - link

    And why is that? Because of the supposed high failure rates? Can you supply any real information about this?

    OCZ has less then 1% failure rate. There may be more then 1% of customers who have "issues" but they aren't related to the drive. User error plays a pretty big role, but of course it MUST be OCZ's fault right?

    Enterprise customers are professionals who know how to install serious hardware like this. And if they don't? OCZ will help install it for them on site. That's what enterprise companies do!

    Reply
  • Troff - Tuesday, September 27, 2011 - link

    I don't believe that 1% number for a second. First of all, I read some return stats from a store that listed the RETURN rate at just below 3%. Secondly, I know of 5 very different systems with Vertex 3 drives in them. All 5 have recurring lockups/BSODs. The people who built and run these systems write their own filesystems. They are extremely knowledgeable. If they can't make them run properly, they are not fit to run outside of a lab environment.

    That said, I suspect it's as much Sandforce that's the problem as it is OCZ.
    Reply
  • josephjpeters - Wednesday, September 28, 2011 - link

    I think it's an Intel problem. But NooOoOo... it can't be an Intel problem... Reply
  • geddarkstorm - Wednesday, September 28, 2011 - link

    From all the data I've been seeing, it seems to be a SATA III issue, and an issue with motherboards not being reading for such high volumes of data flow. Mechanical drives can get no where near SSD speeds, and I don't think manufacturers were really expecting how fast they'd go on SATA III (almost pegging it out at times, and it's brand new!). Reply
  • josephjpeters - Wednesday, September 28, 2011 - link

    Exactly. It's not an OCZ issue, it's the motherboard. When will someone step in and take the blame? Reply
  • Beenthere - Tuesday, September 27, 2011 - link

    SSDs appear to be an on-the-job learning program for SSD manufacturers with all the issues that currently exist.

    I do not however believe they are selling SSDs at low margins.

    Enterprise won't use SSDs yet for the same reason informed consumers won't use them - they have serious reliability and compatibility issues. Unless you can afford lost data and a hosed PC, SSDs are not even an option at this point in time. Maybe in a couple more years they will sort out the problems that should have resolved long agao?
    Reply
  • dave1231 - Tuesday, September 27, 2011 - link

    I wonder really how much a consumer SSD costs to produce. Saying that slim margins will force companies out of business if there's a big markup on a 128GB is not true. These same drives were $100s of dollars last year and probably still aren't good value today. Unless you're saying consumers are waiting for the .50c/GB drive. Reply
  • josephjpeters - Tuesday, September 27, 2011 - link

    It's roughly 20% margins and the price of an SSD is directly related to the cost of Flash. Owning the controller IP is key in maintaining solid margins.

    Enterprise drives will drive flash demand which will lead to economies of scale that result in cheaper Flash prices and consequently cheaper consumer SSD's.
    Reply
  • ChristophWeber - Tuesday, September 27, 2011 - link

    Anand wrote: "I've often heard that in the enterprise world SSDs just aren't used unless the data is on a live mechanical disk backup somewhere. Players in the enterprise space just don't seem to have the confidence in SSDs yet."

    I use an SSD in an enterprise environment, a first gen Sandforce model from OWC. I do trust it with my main workload - database and web server in this case, but of course it is still backed up to mirrored hard drives nightly, just in case.

    I'd have no qualms deploying a Z-Drive R4 in one of our HPC clusters, but it'd be an RM88 model with capacitors, and I'd still run the nightly rsync to a large RAID unit. Now if someone would finally signal they want to spend another $100k on a cluster, and I'll spec a nice SSD solution for primary storage.
    Reply
  • nytopcat98367 - Tuesday, September 27, 2011 - link

    is it bootable? can it b used 4 a desktop too? Reply

Log in

Don't have an account? Sign up now