Yeah a ramdisk would probably work wonders. It's a shame Intel optane didn't became a standard, those type of workflows would be amazing for it.
replyYa know, here on the local market there are a bunch of optanes hanging around, I'll try to manage one to check if there's any improvement
replyOptanes will be good for latency, but not so much for BW which seems to be your major bottleneck if I'm not mistaken?
replyyeah, the mobo upgrade is something I gotta do anyway, so I'll cover that up more or less, the optane is something I didn't thought about
replydeleted
replyAhhh damn it. Intel! Come back!
replyThis is exactly what I was wondering
replyI gave a talk a few years ago at dask summit (conf?) on making the stars align with dask-cudf here. We were helping a customer accelerate log analytics by proving out our stack for nodes that look roughly like: parallel ssd storage arrays (30 x 3 GB/s?) -> GPUDirect Storage -> 4 x 30 GB/s PCIe (?) -> 8 x A100 GPUs, something like that. It'd be cool to see the same thing now in the LLM world, such as a multi-GPU MoE, or even a single-GPU one for that matter!
Isn't m.2 storage but DRAM - hopefully, meaning NVMe/PCIe not SATA speed - already exists as Compute Express Link (CXL), just not in this specific m.2 form factor? If only RAM wasn't silly expensive right now, one could use 31GB/s of additional bandwidth per NVMe connector.
replyThe marvel cxl 2.0 ddr4 card Serve the Home used for kvcache speed ups. And I am personally looking forward to cxl 3 and memory coherence across my system builds.
replyhttps://www.servethehome.com/hyper-scalers-are-using-cxl-to-...