Performance Consistency

Performance consistency tells us a lot about the architecture of these SSDs and how they handle internal defragmentation. The reason we don’t have consistent IO latency with SSD is because inevitably all controllers have to do some amount of defragmentation or garbage collection in order to continue operating at high speeds. When and how an SSD decides to run its defrag or cleanup routines directly impacts the user experience as inconsistent performance results in application slowdowns.

To test IO consistency, we fill a secure erased SSD with sequential data to ensure that all user accessible LBAs have data associated with them. Next we kick off a 4KB random write workload across all LBAs at a queue depth of 32 using incompressible data. The test is run for just over half an hour and we record instantaneous IOPS every second.

We are also testing drives with added over-provisioning by limiting the LBA range. This gives us a look into the drive’s behavior with varying levels of empty space, which is frankly a more realistic approach for client workloads.

Each of the three graphs has its own purpose. The first one is of the whole duration of the test in log scale. The second and third one zoom into the beginning of steady-state operation (t=1400s) but on different scales: the second one uses log scale for easy comparison whereas the third one uses linear scale for better visualization of differences between drives. Click the buttons below each graph to switch the source data.

For more detailed description of the test and why performance consistency matters, read our original Intel SSD DC S3700 article.

  SanDisk Extreme Pro SanDisk Extreme II Intel SSD 730 Intel SSD 530 OCZ Vector 150
Default
25% Spare Area

Similar to the Extreme II, the IO consistency is just awesome. SanDisk's firmware design is unique in the sense that instead of pushing high IOPS at the beginning, the performance drops close to 10K IOPS at first and then rises to over 50K and stays there for a period of time. The higher the capacity, the longer the high IOPS period: the 960GB Extreme Pro takes ~800 seconds before the IOPS drops to 10K (i.e. the drive reaches steady-state). I do not know why SanDisk's behavior is so different (maybe it has something to do with nCache?) but it definitely works well. Furthermore, SanDisk seems to be the only manufacturer that has really nailed IO consistency with a Marvell controller because Crucial/Micron and Plextor have had some difficulties and their performance is not even close to SanDisk.

However, I would not say that the Extreme Pro is unique. Both Intel SSD 730 and OCZ Vector 150 provide the same or even better performance at steady-state, and with added over-provisioning the difference is even more significant. That is not to say that the Extreme Pro is inconsistent, not at all, but for a pure 4KB random write workload there are drives that offer (slightly) better performance.

  SanDisk Extreme Pro SanDisk Extreme II Intel SSD 730 Intel SSD 530 OCZ Vector 150
Default
25% Spare Area

 

  SanDisk Extreme Pro SanDisk Extreme II Intel SSD 730 Intel SSD 530 OCZ Vector 150
Default
25% Spare Area

 

TRIM Validation

To test TRIM, I filled the drive with sequential data and proceeded with 60 minutes of 4KB random writes at queue depth of 32. I measured performance with HD Tach after issuing a single TRIM pass to the drive.

TRIM works for sure as the write speed is at steady 400MB/s.

.

Introduction, The Drives & The Test AnandTech Storage Bench 2013
Comments Locked

85 Comments

View All Comments

  • ninjag - Thursday, June 19, 2014 - link

    Where can I find these 10000% washing machines? I have been tracking this market for years, and I am so tired of incremental 5% gains on washing machine silicon. WE WANT INNOVATION!!!
  • brucek2 - Tuesday, June 17, 2014 - link

    Its discussions like this that make it hard for me to get a feel for whether it would matter to me or not.

    On the one hand, 162% sounds plenty substantial. I do not want to spend 162% more immersion-breaking time twiddling my thumbs waiting for the next level to load, let alone 162% more time to get through a compile so I can meet my next deadline.

    On the other hand, waiting 1/1,000th of a second does not actually sound any faster to me than waiting 1/4,000th of a second (these are the "average service times" in the benchmark), because I know my personal threshold for perceiving laggy response kicks in at around 0.1 seconds.

    So where does this leave me? Not sure really but in the case of doubt it probably goes to spend more. The extra $100-$200 of potential hardware savings would be eaten up quickly in my time to research it much further, to end up ordering & installing & migrating to another drive sooner than I otherwise would have, or if I really did end up twiddling my thumbs more.

    But never fear MyrddinE, I'm still not buying the 99.8% pure unicorn dust super connect audio cables.
  • mickulty - Tuesday, June 17, 2014 - link

    There are plenty of use cases where every little helps - a 5% overclock, assuming linear scaling, would shave ~8.5 minutes off a 3 hour video transcoding task for example.
  • UltraTech79 - Sunday, June 29, 2014 - link

    See but actual users do not do "blind tests" they do actual tests for their servers to see what is better. You dont need to feel the difference between 5 and 10 to know what number is higher.
  • n13L5 - Thursday, December 3, 2015 - link

    In many cases, you are correct. But if you were just teetering near sufficient performance, a 5% performance increase to stay solidly just above 30 fps, the difference will suddenly become very obvious to perceive.
  • UltraTech79 - Sunday, June 29, 2014 - link

    Yes. No one would notice...

    ...Except the sort of people that would read these kinds of articles. Your logic is a failure.
  • vaayu64 - Monday, June 16, 2014 - link

    Thanks for a nice review. Do you know if Sandisk is going to release an mSata version of this ?
    Sadly, there isn't a good msata ssd > 240 GB capacity and with good performance consistency in the market right now....
    Regards
  • 457R4LDR34DKN07 - Monday, June 16, 2014 - link

    I know what you mean because when I built my mITX I think the only drive I could find was the ocz nocti 128GB. If I was going to upgrade it I would go with a Samsung 840 evo but mSATA isn't worth investing into at this point with M.2 arrival IMO.
  • vaayu64 - Monday, June 16, 2014 - link

    The 840 evo msata is without doubt a very nice ssd, but TRIM is not supported on that one.
  • ijozic - Tuesday, June 17, 2014 - link

    Where did you get that (mis)information from?

Log in

Don't have an account? Sign up now