TL;DR: So far, PrimoCache works amazingly on Benchmarks (Blackmagic Disk Test, CystalDiskMark), but actual file copies yield zero or worse performance. I've tried numerous configurations but I think "I'm doing it wrong".
![Smile :)](./images/smilies/icon_e_smile.gif)
Setup:
AMD Threadripper 3690X + 256GB RAM
Asus Strix TRX40-E Gaming (64 Usable Gen4 PCIe Lanes)
2 x 64TB RAID6 Arrays (Areca 1882i)
7 x 2TB NVMe (Gen4 980 Pros, 4x Asus Hyper M.2 Gen4, 3x MB).
I started with the configs from this thread (viewtopic.php?t=4639), specifically setting up the NVMes as L2 Cache specifically to increase write speeds. When I use benchmark apps, I see the acceleration on the spinning drives as if they were the NVMes, which is awesome. But Windows file copy, Xcopy, Robocopy, and TerraCopy all had equal or worse performance (without PrimoCache 1-2 Gb/s, with 1-.75 Gb/s).
I've tried with L1 (up to 224GB), without L1, Defer write on/off, and tried configuring the NVMes as L2 various ways... Individually, by controller, BIOS RAID0, Windows Stripe, checked individual component performances, fresh operating system install (Windows 11), latest drivers, ensured enough PCIe lanes, moved slots... just about everything I could think of.
In all cases I see the Cache Hit Rate happening but it's like it performs as if not doing anything. I feel like I'm missing something. Could this be because I am using a trial version?
Thanks in advance for the help!