0

Just got this Team Force Cardea Zero NVMe PCIe Gen4 drive. It's performance isn't what was promised.

They claim write speeds of 4,400 but I am only seeing 1,500

$ dd if=/dev/zero of=/tmp/test2.img bs=1G count=10 oflag=dsync
10+0 records in
10+0 records out
10737418240 bytes (11 GB, 10 GiB) copied, 7.27396 s, 1.5 GB/s

Are there updates I need to do, or settings I need to change?

3
  • 1
    In the specs: "[1] This product is compatible with Intel and AMD platforms, and the performance result is tested on a motherboard that supports PCIe 4.0 interface by the T-FORCE internal laboratory. The actual speed may vary depending on the software and hardware conditions of the platform." We don't know if there is another bottleneck ...
    – sudodus
    Commented Jan 10, 2020 at 16:51
  • 2
    Only a few new motherboards have PCI Express 4.0 M.2 slot. Does yours? phoronix.com/…
    – oldfred
    Commented Jan 10, 2020 at 17:01
  • Yes, ROG Zenith II Extreme has PCIe4 M.2 slots. Commented Jan 10, 2020 at 18:06

3 Answers 3

3

Plain wrong

$ dd if=/dev/zero of=/tmp/test2.img bs=1G count=10 oflag=dsync

Reason, for one, is that using dd for benchmarking is kind of impossible due to its block size specification need, largely.


I do not use dd ever (again), only in necessary cases, which are rare luckily.


What I recommend to you is installing and running pv (man page).


pv Benchmark suggestion

Example (average rate inclusive test):

pv -Wptearb < /dev/zero > ~/nvme-speed-test
1

Run lspci -vv to see how many lanes and what version of the PCI-Express protocol your SSD is using.

Also bs=1G in dd is/was known to be slow, this might be a bottleneck.

0

I'd suggest to find some reviews of this device online. There are some reputable tech sites (e.g. tomshardware) that produce good quality reviews with tests to see if the product lives up to specification. You can also find interesting and informative reviews on youtube.

You'll find that often specifications show a very bright perspective on the actual performance of the device. This seems fairly common with Flash drives, that are often known to slow down quite badly when put under strain for a longer period of time. One of the possible causes is that the device itself is very slow, but the manufacturer has provided a fast buffer. Once the buffer is full, the operation will basically stall as the device itself is pushed to it's (rather awful) limits trying to catch up.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .