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.

ATSB - Light (Data Rate)

The Crucial MX300 performs very well on the Light test with an average data rate that the MX500 cannot quite match, but the MX500 doesn't lose as much performance when the test is run on a full drive.

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

The average and 99th percentile latency scores of the Crucial MX500 are largely unremarkable, though the 99th percentile latency is near the high end of the normal range. The MX500 is a substantial improvement over the MX300 when it comes to full-drive performance.

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

The average read latency of the Crucial MX500 on the Light test is close to the Samsung 850 PRO and EVO when the test is run on an empty drive, but is merely average when the drives are full. The average write latency is a bit below average in both cases, but the full-drive penalty is much reduced compared to the MX300.

ATSB - Light (99th Percentile Read Latency)ATSB - Light (99th Percentile Write Latency)

As with the average read and write latency scores, the 99th percentile read and write scores fall within the normal range. The 99th percentile read latency is a bit better than average while the 99th percentile write latency is worse than most drives, but the MX500 isn't an outlier in either direction.

ATSB - Light (Power)

The power consumption of the MX500 ranks a bit better on the Light test than it did on the Heavy test. The Crucial MX300 is still substantially better, and the slow but DRAMless Toshiba TR200 holds on to a comfortable lead. The Samsung 850 PRO and EVO are in last place.

AnandTech Storage Bench - Heavy Random Performance
POST A COMMENT

90 Comments

View All Comments

  • mode_13h - Tuesday, December 19, 2017 - link

    Actually, you don't need it to upgrade the firmware. But it also does things like over-provisioning and secure-erase. And it can tell you roughly how much write endurance is remaining. Reply
  • ddrіver - Tuesday, December 19, 2017 - link

    Overprovisioning, that very advanced function that translates to shrinking a partition and leaving some free space somewhere at the end of the drive using nothing but Windows tools... There is no killer function in these SSD tools. They make FW updates a lot easier for regular people. I don't want to make boot disks and type in my updates. Just click click form the GUI and I'm done. Maybe something like Samsung's RAPID needs the software but other than that you can do it yourself. Reply
  • jabber - Tuesday, December 19, 2017 - link

    Yeah I just use Disk Management or Partition Wizard to leave 2-5GB (depending on size) free on all my SSDs. Reply
  • mode_13h - Wednesday, December 20, 2017 - link

    Good luck with that. First, it's not much (I usually OP by anywhere from 6.25% to 25% - if you search SSD reviews on this site, you'll see performance benefits even up to 25%).

    Second, it definitely won't work for all vendors. The SSD firmware needs to know that the unpartitioned space doesn't hold user data, and they don't all derive this information by simply looking for a partition table and parsing that, as you seem to believe.
    Reply
  • ddrіver - Thursday, December 21, 2017 - link

    Any link for that? The point of overprovisioning is to have space to swap data and do internal reorganization even when the drive is full, for wear leveling. Since most drives support TRIM and you can trigger it manually it's impossible to assume there will be data there. It's like the SSD should stop with the wear leveling because I might have some data in that free space it's using anyway.

    The ONLY difference between normal free space and OP is that OP will be there even when you regular partition is full. Wear leveling and data reorganization works even with 0 OP when there actually is user data everywhere. It just takes longer.

    The second you create an empty chunk of disk space it will be TRIMmed anyway and it becomes really free space. It can even be a partition that you never write to. Windows will TRIM it regardless and the SSD will know the LBAs are empty. No flag needed. But I'd love to see some docs from recent times to say otherwise.
    Reply
  • mode_13h - Friday, December 22, 2017 - link

    I don't know why you think anyone is TRIMming unpartitioned space, but it's a bad assumption. Reply
  • mode_13h - Wednesday, December 20, 2017 - link

    Their documentation seems to suggest it writes something to flag the unpartitioned space as useable for over-provisioning. I don't know how you can easily prove that simply leaving unpartitioned space is equivalent.

    With certain other vendors, I've seen official statements that simply leaving unpartitioned space is not sufficient for overprovisioning.
    Reply
  • npz - Tuesday, December 19, 2017 - link

    In addition to secure erase, you also need it for Opal / eDrive. Reply
  • Wolfpup - Tuesday, December 19, 2017 - link

    Dumb question, but what do you need to use it for? I've never used it, that I can remember, and I've got a couple of Crucial drives. I don't have Java on any of my personal systems either. Reply
  • mikato - Friday, January 19, 2018 - link

    Agree. I don’t see how Java is a problem. Reply

Log in

Don't have an account? Sign up now