You are not logged in.
I would like to test one of my USB drive for speed, is there any way I can do this?
Offline
If you ask about that, then the sequential write speed is about 5MB/s.
If you had anything better, you would know. :D The manufacturer would advertise that like mad and you would consciously spend noticeable amount of money on getting better characteristics. Otherwise you just got the baseline. That is 3–10MB/s, typically about 5MB/s.
In case you want to verify if the manufacturer is not lying to you or test an unknown device:
head -c 1GB /dev/urandom >/tmp/testfile
time bash -c 'cd /path/to/mounted/usb && mv /tmp/testfile . && sync -f testfile'
Divide 1GB by time reported as “real” to get sequential write speed to a filesystem on a USB. That is the speed you may hope to get in actual use.
Last edited by mpan (2022-06-25 05:40:49)
Sometimes I seem a bit harsh — don’t get offended too easily!
Offline
I found that hdparm can do that
$ hdparm -t /dev/sda
Timing buffered disk reads: 60 MB in 3.05 seconds = 19.68 MB/sec
While my others stick are twice faster:
Timing buffered disk reads: 118 MB in 3.04 seconds = 38.77 MB/sec
Last edited by sak18ra (2022-06-26 00:58:16)
Offline
Note you only reported read speeds, is that all you care about? If so, something's still fishy. The "others" are faster, but those speeds don't seem to match usb specs. The first one is within USB 2.0 range (and on the better end of it), but the "others" are faster than USB 2.0 should be capable of but also way too slow to be USB 3.0.
"UNIX is simple and coherent" - Dennis Ritchie; "GNU's Not Unix" - Richard Stallman
Offline
40 MB/s seems within the limits of USB 2.0, as reported in this test. But, as Trilby has noted, that’s the raw sequential read speed. Expected to be higher than the write speed even by an order of magnitude, but showing only half of the picture at best. There is only a handful of applications, where this parameter matters.⁽¹⁾ What limits you in everyday use is how fast you can write files to the filesystem and the pull then thumb drive out. And that’s what the proposed shell code tests: every piece of it is there for a reason and the choice of tools is deliberate.
If you wish, there are also specialized tools offering such benchmarks. extra/gnome-disks-utility can test random access reads and writes (the test is destructive).
____
⁽¹⁾ A boot medium for a live distro, to give one example.
Last edited by mpan (2022-06-26 09:08:53)
Sometimes I seem a bit harsh — don’t get offended too easily!
Offline