Has anyone ever heard of companies actually using Ponte Vecchio?
Benchmarks are excellent... but every git project I have read is trained on a Nvidia x100 cluster. And there were maybe 2 repositories than mentioned ROCm/CDNA on a single line in passing.
I cant recall ever seeing a cloud instance with it.
OpenVINO is also stupidly easy to use in PyTorch. Stable Diffusion (for instance) works with a single import and changing the device from "gpu" to "xpu." This is not my experience with other accelerators.
But alas if researchers cant peel themselves away from Nvidia, maybe Intel is smart to drop it and focus on something they can push as a CPU host, which might actually get researchers to use it.
So far all the Ponte Vecchios are going to Argonne for Aurora. Technically it was launched but I can't find the price and I don't think it's actually being sold.
Xeon Phi could act as a host and people still didn't use it.
Benchmarks are excellent... but every git project I have read is trained on a Nvidia x100 cluster. And there were maybe 2 repositories than mentioned ROCm/CDNA on a single line in passing.
I cant recall ever seeing a cloud instance with it.
OpenVINO is also stupidly easy to use in PyTorch. Stable Diffusion (for instance) works with a single import and changing the device from "gpu" to "xpu." This is not my experience with other accelerators.
But alas if researchers cant peel themselves away from Nvidia, maybe Intel is smart to drop it and focus on something they can push as a CPU host, which might actually get researchers to use it.