Driver performance - CrystalDisk Mark

shameful bump

Saged

>>56491964
why? :(

>>56491702
Turn RAPID off you useless waste of air.

>>56492192
why?

>>56492320
Because best case scenario it does nothing, and worst case scenario it makes things worse.

File: Screenshot (4).png (887KB, 1920x1080px) Image search: [Google]

887KB, 1920x1080px

>>56492332
These are the results after disabling RAPID mode

>>56492098
You can google these things yourself. Then again it is one of the better threads up now so
Bump

>>56492472
And there you have the real results. Good job.

>>56492539
noone is posting results though :(

>>56494163
Because all /g/ cares about is GPUs and applel

>>56494664
frogposters

>>56494751
Sadly

File: crystaldiskshit.png (12KB, 398x337px) Image search: [Google]

12KB, 398x337px

>>56491702
2x 2TB WD Red WD20EFRX
2x 3TB HGST UltraStar 7K4000

RAID10 (ZFS)

File: crystaldiskshit_32gb.png (13KB, 398x337px) Image search: [Google]

13KB, 398x337px

>>56495015
And here's what happens when I bump up the working set beyond what my system is designed for

>>56495015
>>56495146
Also, it goes without saying that this is not a very realistic benchmark for most use cases because the access pattern is too spikey - this data does not survive long enough to enter L2ARC, so my SSD cache devices are not being used at all.

>>56495015
>>56495146
Are those the performances of the SSD?

>>56495623
No. The SSD was basically not used at all during this test, it was almost entirely ARC for the first test. The second test got more complicated, because it started out with HDD speed and then started moving data into ARC as it got accessed.

Benchmarking ZFS is very nontrivial because there are so many different variables you need to isolate. I posted these results mostly for fun, not to be taken too seriously.