SSD read/write speed discrepancy between CrystalDiskMark 8 and AS SSD

Yesterday i returned my main SSD, an Inland Premium 1TB to my motherboard after running a SATA drive for a few days for testing.
Anyway, i decided to benchmark the drive since i haven’t done it in a year or more, i downloaded CrystalDiskMark 8 and went to town.
Read speeds are correct, but write speeds concerned me a bit, it was way lower than last year, i tried changing modes in Crystal to make sure, but it either gave me the same or worse results.
I downloaded AS SSD, which i had never used before and tested again, now that one gave me a much more similar result in comparison with last year’s test, so, what’s the deal with the result discrepancy?
I also noticed, comparing CrystalDiskMark 8 and CristalDiskMark 6 screenshoots, is that the benchmarks have different names, therefore different functions, so maybe the result is correct and i’m just being paranoid?
Here are the benchmarks:

Last year (CrystalDiskMark 6):

Yesterday (CrystalDiskMark 8 + AS SSD):

Get third one, go with majority :wink:

Also disk benchmarks are usually more finicky than gpu/cpu stuff. Depends on many variables. And with SSD’s situation got even more complicated with complicated controlers, dram, variable cell levels etc. You basically benchmarking another computer.

1 Like

Is there a 3rd one you’d suggest me checking out?

For Windows? ATTO if they still around.
bonnie++, hdparm or even dd for Linux.

TPup has download for it, you may even be able to compare results to other benchamrks if you find your model:

1 Like

Thanks! I’ll check it out when i get home.

Got some more, ATTO and CrystalDiskMark6.
Now i’m downloading Karsperky just to be sure, haven’t checked for viruses in this machine in about as long as i didn’t do any benchmarks.

Kaspersky results are in, nothing, my system is as clean as ever.
I also found out i’m not the only one with this issue, and i’m very late discovering it:

Also updated the firmware on the drive from ECFM 12.2 to 12.3 and there was no difference.

@Automobili3XF Are you aware of the need to “erase” flash memory after use before writing to it again, and the fact that erasure is much slower than either reading or writing? Perhaps last year your SSD had a lot of never-written blocks, but this year the blocks have all been used, so they must be erased before being written?

There is a process called TRIM by which free-but-previously-used blocks are erased so they can be written quickly when the need arises. Do you know whether this is in use for your drive? It takes some setup in Linux; I don’t know about Windows.

Yes, I’m aware, on Windows 10 it’s set up to be done every week.
I was wondering if that could be the case, i checked it multiple times and even tried to manually TRIM the drive with powershell commands, but there was no improvement after manual or automatic TRIM.

This topic was automatically closed 273 days after the last reply. New replies are no longer allowed.