You are not logged in.
Hello,
I have an Intel 750 series 1.2TB PCIe SSD. This drive works fine in Windows 10 (x64) according to speed specifications by Intel, but on Arch Linux (x64, 4.2.5-1) i am getting different results.
According to Intel and benchmarks done on Windows i should be getting a sequential read speed of around 2500 MB/s. I am only getting 1200-1300 MB/s on Arch Linux. Write speed are the same, but correct according to specifications.
I have tested the drive the following way:
[root@citadel ~]# timeout -s SIGINT 30s dd if=/dev/nvme0n1 bs=4K of=/dev/zero
4850017+0 records in
4850016+0 records out
19865665536 bytes (20 GB) copied, 29.9986 s, 662 MB/s[root@citadel ~]# timeout -s SIGINT 30s dd if=/dev/nvme0n1 bs=64K of=/dev/zero
608805+0 records in
608804+0 records out
39898578944 bytes (40 GB) copied, 29.999 s, 1.3 GB/s[root@citadel ~]# timeout -s SIGINT 30s dd if=/dev/nvme0n1 bs=1M of=/dev/zero
38250+0 records in
38249+0 records out
40106983424 bytes (40 GB) copied, 30 s, 1.3 GB/s[root@citadel ~]# timeout -s SIGINT 30s dd if=/dev/nvme0n1 bs=4M of=/dev/zero
9516+0 records in
9515+0 records out
39908802560 bytes (40 GB) copied, 30.0006 s, 1.3 GB/sI clear the cache before each test with:
sync && echo 3 > /proc/sys/vm/drop_cachesIm getting similar results using an ext4 filesystem on the drive and testing with actual files.
My system:
- ASUS RAMPAGE IV EXTREME
- Intel i7 4960x
- 32GB RAM
How do i get the sequential read speeds according to the specifications?
flnhst
Offline