Performance Consistency

Performance consistency tells us a lot about the architecture of these SSDs and how they handle internal fragmentation. The reason we do not have consistent IO latency with SSDs 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 (Logical Block Addresses) 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 dropdown selections 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.

Transcend SSD370 256GB
Default
25% Over-Provisioning

Despite the custom Transcend firmware, performance consistency is an exact match with ADATA's SP610. I'm suspecting that the reason for low steady-state performance might be the hardware because the SM2246EN is a single-core design. Most controller designs today are multicore because today's NAND requires a lot of management and with multiple cores the NAND management can be dedicated to one or more cores, which leaves the rest of the cores available for host IO processing. In Silicon Motion's case, the one core has to take care of everything from host IOs to NAND management, which translates to lower overall performance as the controller can't keep up with everything that needs to be done.

Transcend SSD370 256GB
Default
25% Over-Provisioning

 

Transcend SSD370 256GB
Default
25% Over-Provisioning


TRIM Validation

To test TRIM, I filled a 128GB SSD370 with sequential 128KB data and proceeded with a 30-minute random 4KB write (QD32) workload to put the drive into steady-state. After that I TRIM'ed the drive by issuing a quick format in Windows and ran HD Tach to produce the graph below.

And TRIM works as expected.

Introduction, The Drive & The Test AnandTech Storage Bench 2013
POST A COMMENT

44 Comments

View All Comments

  • hojnikb - Tuesday, January 27, 2015 - link

    geizhals.at also finds 32GB version for 32€. Although i dont think many people will buy this. Reply
  • Maltz - Tuesday, January 27, 2015 - link

    "The StaticDataRefresh technology monitors the error rates and when a preset threshold value is reached, the data will be rewritten to restore the correct cell charge level. I suspect all SSDs do this because it's vital to ensure the health of old data, but it's the first time I've seen it mentioned in a data sheet."

    I've also believed this for some time. This is a little off-topic, but doesn't this mean that TRIM is more important to drive longevity than is widely believed? Sure, garbage collection and over provisioning can usually maintain a drives' performance levels, but if the drive is re-copying around unused blocks, then this seems like a problem. (I'm looking at you, Apple!)
    Reply
  • Kristian Vättö - Tuesday, January 27, 2015 - link

    Technically yes, because the drive would be rewriting invalid (i.e. deleted) data as well. However, even without TRIM the drive will know what pages are invalid once the OS writes to the same LBAs again. The invalid pages will then be deleted sometime during garbage collection, which will return the drive to "TRIMed" state (i.e. no invalid data).

    It's basically the same with TRIM too because the drive doesn't necessarily erase the data immediately (i.e. data will initially be written to OP space). TRIM merely gives the drive a heads up and the data can be deleted when appropriate, whereas non-TRIM system will give the heads up when there' already new data coming in.
    Reply
  • Gc - Tuesday, January 27, 2015 - link

    | p7: Drive Power Consumption - Random Write
    | ...
    | Transcend SSD370 128GB - 1.90
    | ...
    | Transcend SSD370 512GB - 2.73
    | ...
    | Transcend SSD370 256GB - 3.12

    Anyone have an explanation for how the middle 256GB size used the most power in this test?
    (or was there a typo?)
    Reply
  • IlikeSSD - Wednesday, January 28, 2015 - link

    with new OCZ prices I'd rather go for Arc... http://www.kitguru.net/components/ssd-drives/leo-w... Reply
  • velanapontinha - Wednesday, January 28, 2015 - link

    I've been waiting for this review for months. These drives were being showcased inside a low end server at Computex. I bought for 512GB drives and put them inside a HP server, connected in RAID 5 to a P410i controller.
    After a few months I started experiencing one drive failing every week. It all went downhill, as the failures started ocurring almost every day.
    I removed those drives and they are now working fine inside laptops.
    Transcend support told me they could not provide any support, as these drives were not tested in server environments.

    Although these are working fine and fast in laptops, I was a bit disappointed that they show it running in servers at Computex and then fail to support that same scenario in the real world.
    Reply
  • velanapontinha - Wednesday, January 28, 2015 - link

    Please read "I bought *four 512GB drives".
    Also, I'd like to add that when placed in laptops, the drives reported 100% health as per Transcend's app.
    Reply
  • editorsorgtfo - Thursday, January 29, 2015 - link

    "the SSD370 is also listed at even lower prices on Amazon Prime right now"

    What about regular Amazon? I don't have a Prime account.
    Reply
  • cbjwthwm - Saturday, January 31, 2015 - link

    What firmware was used on the Plextor M6S during testing, or are those old historical results provided for comparison testing? Newer Plextor firmwares (released over the last 6 months) are supposed to have addressed their service time issues, and it would be interesting to see Crucial / Micron's M550 and M600 with their recent (early Jan 2015) firmwares that hopefully address the same issues. Reply
  • Gizbeat - Monday, March 9, 2015 - link

    Thanks for the extensive review. I recently purchased the Kingmax 256GB SME35 which has the same controller and memory. So far the drive has been excellent. Reply

Log in

Don't have an account? Sign up now