Hacker News new | ask | show | jobs
by woodrowbarlow 26 days ago
i wish there was a system like strix halo, but with enough lanes for a dedicated PCIe 5.0 x16 slot so you can have the best of both worlds: large sparse models on CPU with unified memory, dense models on GPU with real tensors and higher bandwidth memory.
2 comments

AMEN! 100% agree.

Plus a reasonably inexpensive super low-latency interconnect.

Maybe this is what Medusa Halo will turn out to be?