First, thank you to Fernando and the other Gurus for keeping this forum and it’s guides up. Although I just recently became a member, it was via the guides, benchmarks, and the like that I decided to join the community. Excellent work, clear and descriptive guides and answers… couldn’t ask for more. On that note, hello!
Now to my storage conundrum.
I’ve run into some rough write speeds on my array. Now, I don’t know whether this is attributable to installing Windows 10 x64 or not but I’m guessing it’s not, it’s something I’m missing configuration wise. After reading Fernando’s guide on RST RAID0 Pre-install drivers with and without software, I decided to see if the driver was the issue. So currently my read / write is @ r: 1536/w: 676. I have no idea why it’s so low, I have done some very extensive testing.
So here’s the current setup:
- 3x Kingston HyperX 3K 240gb SSD in RAID0 (RST Pre-install driver 184.108.40.2069 and Software Installed)
- Motherboard RST = 220.127.116.114
- 2x 1TB WD Black on Separate Marvell Controller
- Windows 10 x64 Fresh Install, blank aside from updates and necessary drivers.
- winsat formal -V ran
Device Manager Setup Configuration (Intel Raid 0 Volume Properties):
- Enable write caching on the device = YES/Checked
- Turn off Windows write-cache buffer flushing on the device = YES/Checked
Intel RST 18.104.22.1689 Management Software Config.:
- Write-cache buffer flushing: Disabled
- Cache mode: Off
I have tested each individual drive on RST drivers going back to the lowest ranked in Fernando’s RAID0 on Z97 guide, each time drive wiped using same testing methodology. The write speed seemingly is STILL Half of the read. This applies even in the current 3x SSD RAID0 setup (Not how typically ran it, usually 2x SSD Raid0 but just to test something, it’s currently configured that way on my 200th re-install :P). So, too summarize, I’m wondering if it’s a configuration issue? Would updating the BIOS RST help any, if at all? And why does CrystalDiskMark set to “test data - default (Random)” display half write speed when testing at 0 fill seems to give somewhat normal readings? I guess maybe I’m benching wrong?
Sorry for the long winded first post!
EDIT: Annnnnd of I course I posted this in the wrong part of the forum?