AnandTech Storage Bench - Light

Our Light storage test has relatively more sequential accesses and lower queue depths than The Destroyer or the Heavy test, and it's by far the shortest test overall. It's based largely on applications that aren't highly dependent on storage performance, so this is a test more of application launch times and file load times. This test can be seen as the sum of all the little delays in daily usage, but with the idle times trimmed to 25ms it takes less than half an hour to run. Details of the Light test can be found here. As with the ATSB Heavy test, this test is run with the drive both freshly erased and empty, and after filling the drive with sequential writes.

The Patriot Hellfire, in blue, is highlighted as an example of a last-generation Phison E7 drive. Although we didn't test it at the time, the MP500 was based on the same controller and memory.

ATSB - Light (Data Rate)

The average data rates from the Corsair Force MP510 show a much larger performance hit for running the Light test on a full drive than we saw with the Heavy test, but in either case the MP510 maintains competitive performance.

ATSB - Light (Average Latency)ATSB - Light (99th Percentile Latency)

The average and 99th percentile latencies from the MP510 are very low, though most of the NVMe drives listed here have sub-millisecond 99th percentile latencies and the differences generally imperceptible for workloads this light.

ATSB - Light (Average Read Latency)ATSB - Light (Average Write Latency)

The MP510 continues to have some of the best write latencies and competitive read latencies, with the write latency in particular showing almost no penalty for running the test on a full drive.

ATSB - Light (99th Percentile Read Latency)ATSB - Light (99th Percentile Write Latency)The 99th percentile write latency of the Corsair MP510 on the Light test is minimal, showing that the entire test operates within the very fast SLC write cache. The read latencies are very good, but don't set any records.

ATSB - Light (Power)

The energy usage by the MP510 during the Light test is a bit better than average but still about 30% higher than the Crucial MX500 mainstream SATA SSD.

AnandTech Storage Bench - Heavy Random Performance
Comments Locked

42 Comments

View All Comments

  • leexgx - Thursday, October 18, 2018 - link

    be nice if they do a renew on it as from unreliable source that did a review (toms hard) seems to find the P1 is only a little faster then a MX500 (yes the P1 its a NVME ssd but that's only good for sequential test it seems)
  • yoyomah20 - Thursday, October 18, 2018 - link

    I've been waiting for this review to come out. I'm excited about what corsair has put out, seems like its a pretty good competetor to 970 EVO and WD Black at a cheaper price point. I've been waiting for a power efficient nvme drive to replace my laptop's stock 128GB sata m.2 drive and I think that this is the one! Too bad it's not available anywhere yet...
  • G3TG0T - Thursday, October 18, 2018 - link

    Somehow the price SHOT up by double...
  • G3TG0T - Thursday, October 18, 2018 - link

    Who would buy that for double the price when you could get an EVO 970??!
  • lilmoe - Thursday, October 18, 2018 - link

    Damn Amazon and their sketchy crap. Go to newegg, the price is slightly up 10% though.
  • Lolimaster - Thursday, October 18, 2018 - link

    The other thing is using Office 365 Home, 6TB for $99 a year.
  • shabby - Thursday, October 18, 2018 - link

    Would be nice if all sizes were tested and not just the fastest, you guys should tell oems to send your all the sizes to test.
  • leexgx - Thursday, October 18, 2018 - link

    i could imagine that would take some time to test them, as i would guess Billy/reviewer runs the tests at least 2-3 times to make sure the results are consistent (not looked at the article yet but i guess it was the 1TB one they reviewed)
  • WatcherCK - Thursday, October 18, 2018 - link

    Do OSS NAS solutions (OMV/FreeNAS/Ubuntu+ZOL...) support fast/slow storage tiers transparently? I guess this would look like monolithic storage with the OS caching higher use files behind the scenes... hmmm, how hard would it be to have a hybrid drive that makes use of TLC/QLC (not in a fast caching scenario but say 512GB of TLC and 4/6/8TB QLC in one enclosure and a controller that can present both storage arrays transperently to the OS, an SSD only version of a fusion drive for example.)

    And agree with other posters about capacity, once 96 layer becomes ubiquitous then SSDs should be able to reach parity with mechanical HDD in terms of density and price as far as non enterprise users are concerned...
  • Wolfclaw - Friday, October 19, 2018 - link

    Not fussed about top end speed, just cheap mass storage in raid or Microsoft Storage, that wipes the floor with HDD's and can satuate a SATA3 interface is more than enough for me.

Log in

Don't have an account? Sign up now