Performance Consistency

We've been looking at performance consistency since the Intel SSD DC S3700 review in late 2012 and it has become one of the cornerstones of our SSD reviews. Back in the days many SSD vendors were only focusing on high peak performance, which unfortunately came at the cost of sustained performance. In other words, the drives would push high IOPS in certain synthetic scenarios to provide nice marketing numbers, but as soon as you pushed the drive for more than a few minutes you could easily run into hiccups caused by poor performance consistency.

Once we started exploring IO consistency, nearly all SSD manufacturers made a move to improve consistency and for the 2015 suite, I haven't made any significant changes to the methodology we use to test IO consistency. The biggest change is the move from VDBench to Iometer 1.1.0 as the benchmarking software and I've also extended the test from 2000 seconds to a full hour to ensure that all drives hit steady-state during the test.

For better readability, I now provide bar graphs with the first one being an average IOPS of the last 400 seconds and the second graph displaying the standard deviation during the same period. Average IOPS provides a quick look into overall performance, but it can easily hide bad consistency, so looking at standard deviation is necessary for a complete look into consistency.

I'm still providing the same scatter graphs too, of course. However, I decided to dump the logarithmic graphs and go linear-only since logarithmic graphs aren't as accurate and can be hard to interpret for those who aren't familiar with them. I provide two graphs: one that includes the whole duration of the test and another that focuses on the last 400 seconds of the test to get a better scope into steady-state performance.

Steady-State 4KB Random Write Performance

TLC SSDs in general don't perform well under sustained random workloads and the Trion 100 is no exception. It seems that with TLC SSDs we have to accept a drop from 5K to 2-3K IOPS, which is still fine for basic client workloads since most SSDs from 3-4 years ago couldn't even match that.

Steady-State 4KB Random Write Consistency

The consistency isn't great either, but it's not particularly poor either when compared agains Silicon Motion drives.

OCZ Trion 100
Default
25% Over-Provisioning

Looking at the performance over time, we can see that the baseline performance hovers at about 1,000 IOPS with frequent peaks occuring at 5K to 10K IOPS. That's actually very similar to the Neutron XT's (Phison S10) graph because the baseline is also 1,000 IOPS, although the peaks are higher and more frequent, but it further reassures that the underlying firmware architecture is similar. Increasing over-provisioning doesn't increase the baseline performance, but it does make peak performance moments more frequent and higher (from 10K IOPS to 25K IOPS). 

OCZ Trion 100
Default
25% Over-Provisioning
Introduction, The Drives & The Test AnandTech Storage Bench - The Destroyer
POST A COMMENT

65 Comments

View All Comments

  • valnar - Friday, July 10, 2015 - link

    19nm process *and* TLC *and* a brand new ECC method? No thanks. Sounds too risky. Reply
  • NvidiaWins - Friday, July 10, 2015 - link

    There's a reason OCZ went bankrupt....their SSD's are HORRIBLE! Reply
  • hojnikb - Saturday, July 11, 2015 - link

    Good thing this is OCZ only by name.
    Read the d*** article.
    Reply
  • ES_Revenge - Sunday, July 12, 2015 - link

    Certainly a fail for OCZ/Toshiba. You'd think they would have "went back to the drawing board" with this instead of "releasing to the market". Ugh.

    But, OCZ does have some very good drives out there. Vector 150/180, Vertex 460/A, Radeon R7 drives are all very good (Vector with some impeccable services times and very fast at regular-people workloads), but in the SSD market competition is pretty stiff. If you can get one of the above OCZ drives at a good/sale price they can be excellent drives for the money. The ARC 100 was somewhat disappointing but not terrible so I don't see them as worth buying unless you're getting them at a really good price. The Trion though, not much else to say but fail.

    Looks very much like this is a Phison-controlled drive and though Phison isn't bad it seems to require very specific NAND/configuration to actually shine. Either the TLC handling of the S10 is just poor or Toshiba's TLC isn't that great, or both/combination.
    Reply
  • doggface - Monday, July 13, 2015 - link

    The first manufacturer who can get a 128/256 GB ssd to be (within a few dollars of) the same price as a 5400rpm 500GB HDD -- Will win at life. Full stop. End of argument. All those rubbish midrange laptops that have everything a person needs except fast storage.... All they need is fast storage and they would be very capable and would catapult any OEM into a leading position in that demographic. If they can get this ssd down to those prices for OEM (I realise this is retail) then they have a winner. As an enthusiast drive... Massive fail. Reply
  • doggface - Monday, July 13, 2015 - link

    Seriously, This is the perfect midrange laptop.

    13.3" display IPS
    i5 or i3 (AMD equiv) or top cherrytrail.
    4gb ram expandable.
    128GB ssd
    WiFi-n 5ghz.
    Usb3/HDMI, camera etc
    A decent keyboard
    Clean looks.

    If you can do that for $5-600 USD.
    Millions of sales.

    (U can do a better version with a FHD screen, faster WiFi/cpu/dGPU, bigger capacity SSD.)
    Reply
  • Pessimism - Monday, July 13, 2015 - link

    -Marketing from a dead brand with a bad reputation for reliability... Check!

    -Design from a stubborn, tight lipped old school Japanese megacorp who refuses to disclose technical details... Check!

    -Controller sourced from a bottom-feeding Taiwanese manufacturer accustomed to designing barely functional solutions at minimal cost for OEMs... Check!

    Recipe for a winner here!
    Reply
  • LazloPanaflex - Tuesday, July 21, 2015 - link

    LOL, well said! Reply
  • abufrejoval - Monday, July 13, 2015 - link

    Clearly this drive wasn't designed for ultimate performance, but as I replace more and more of my HDDs with SSDs I begin to worry more about the reliability and endurance of the storage, than the performance:

    How sure can I be, that after five years of mixed usage, some intensive I/O, some stuff (e.g. baby pictures) only ever written once, the SSD perhaps lying around for two years in a safe or another having suffered through years in a quiet but rather warm passive PC every bit of data will come off exactly as it was written? Add automobile scenarios or others with lots of electronic noise to make things harder.

    After all storage isn't just about pounding these things with random bits and see how long they last, it's about delivering on a promise to return everything as it was written a millisecond or years ago.

    Anandtech can't really test that, even if they do an incredible job at testing what can be tested within such a short time and a limited budget.

    But perhaps it's exactly in that area, the reliability under almost all circumstances, where Toshiba shines and others less so?

    So far I never really worried, because my SSDs mostly contained cache data which was quickly forwarded to a RAID of trusted rotating rusties or had a backup copy on them: The overall risks and exposure times were low and seemed well managed with the tests done.

    But as the rusties finally spin down and SSDs become the prime medium for all my data, perspective and requirements change.

    For example the data decay discussion, where retention time depends and varies widely with write temperature and storage temperature.

    If you retire an older SSD (as you might with HDDs) at 80% overwrite capacity used to become a backup device, write data with the drive barely warming beyond room temperature and then put it away in the cellar, dry and above freezing, and then plug it in a year or two later: Will you still get your data back?

    Since the drive doesn't have a clock: How does it even know that it should start checking for and compensating bit rot?

    What does that mean for your backup strategy: Should you fully rewrite the data every couple of months? Should pre-warm the drive before rewriting to ensure optimal retention?

    I can't believe Toshiba was simply unable to design a product which was fast and economic. I'd like to believe that perhaps we are missing their design goal.

    Easy to do, when they are keeping so quiet :-)
    Reply
  • rocketman122 - Tuesday, July 14, 2015 - link

    I wouldnt touch ocz if they gave it to me for free. such garbage and bad ethics as a company. didnt they file for bankruptcy? we dont need crap companies that make drives where people store important data on them. after selling the core series years back with the jmicron chips and knowing they were defective I have no respect for them. I tell anyone that asks about components NEVER to buy OCZ. Reply

Log in

Don't have an account? Sign up now