Re:Intel announces Arc B-series "Battlemage" discrete graphi...
by jsheard on Tuesday December 03, 2024 @04:02PM
(#42311378)
The question is whether there's enough overall demand for a GPU architecture with 4x the VRAM of a 5090 but only about 1/3rd of the bandwidth. At that point it would only really be good for AI inferencing, so why not make specialized inferencing silicon instead?
[ Parent
| on HN ]
Re:Intel announces Arc B-series "Battlemage" discrete graphi...
by mandelken on Tuesday December 03, 2024 @05:38PM
(#42312407)
I genuinely wonder why no one is doing this? Why can't I buy this specialized AI inference silicon with plenty of VRAM?
Re:Intel announces Arc B-series "Battlemage" discrete graphi...
by ahakki on Tuesday December 03, 2024 @06:31PM
(#42312899)
I guess that would be an NPU combined with LPDDR. Basically any Windows Copilot Plus approved device.
Re:Intel announces Arc B-series "Battlemage" discrete graphi...
by ryao on Wednesday December 04, 2024 @12:24AM
(#42314676)
Intel and Qualcomm are doing this, although Intel uses HBM and their hardware is designed to do both inference and training while Qualcomm uses more conventional memory and their hardware is only designed do inference:
https://www.intel.com/content/www/us/en/products/details/pro...
https://www.qualcomm.com/news/onq/2023/11/introducing-qualco...
They did not put it into the PC parts supply chain for reasons known only to them. That said, it would be awesome if Intel made high memory variants of their Arc graphics cards for sale through the PC parts supply chains.