Linux rock64 4.4.70-gd9f8397-dirty #4 SMP Thu Jun 29 21:45:40 UTC 2017 aarch64 aarch64 aarch64 GNU/Linux
Corsrair Flash Voyager GT3.0 128GB (ext4)
For comparison on my Thinkpad W520 (i7@3.3Ghz) with linux-deepblack 4.11.5-200.fc25.x86_64
So considering that this kernel on Rock64 is pretty new this does not look bad at all...
Edit: The Thinkpad values have some weird bumps in them, maybe the file size is too small and testing some caching or buffer here instead of the physical device, going to rerun with large filesize...
Edit: re ran both tests with larger file size and the results are more consistent now...
Corsrair Flash Voyager GT3.0 128GB (ext4)
Code:
random random
kB reclen write rewrite read reread read write
409600 4 1640 1467 14241 15181 4174 571
409600 16 5769 6545 43072 48283 19409 2867
409600 512 24039 26066 135432 134970 112677 12135
409600 1024 24033 27872 132993 133170 133392 13713
409600 16384 27784 28047 135172 135434 137017 13825
For comparison on my Thinkpad W520 (i7@3.3Ghz) with linux-deepblack 4.11.5-200.fc25.x86_64
Code:
random random
KB reclen write rewrite read reread read write
409600 4 1738 1689 22762 22883 4953 724
409600 16 5841 6673 70025 70230 24736 2878
409600 512 62620 76978 212627 211742 161889 24058
409600 1024 33786 53683 202181 200167 175399 18183
409600 16384 72790 70291 218358 218548 214972 23091
So considering that this kernel on Rock64 is pretty new this does not look bad at all...
Edit: The Thinkpad values have some weird bumps in them, maybe the file size is too small and testing some caching or buffer here instead of the physical device, going to rerun with large filesize...
Edit: re ran both tests with larger file size and the results are more consistent now...
Come have a chat in the Pine IRC channel >>