Re:Making AMD GPUs competitive for LLM inference
by derektank on Tuesday December 24, 2024 @12:40AM
(#42499937)
Wait, are they losing money on every one in the sense that they haven't broken even on research and development yet? Or in the sense that they cost more to manufacture than they're sold at? Because one is much worse than the other.
[ Parent
| on HN ]
Re:Making AMD GPUs competitive for LLM inference
by rockskon on Tuesday December 24, 2024 @02:52AM
(#42500364)
They're trying to unseat Radeon as the budget card. That means making a more enticing offer than AMD for a temporary period of time.
Re:Making AMD GPUs competitive for LLM inference
by ryao on Tuesday December 24, 2024 @08:45PM
(#42506275)
The former is likely true, but the latter is not:
https://news.ycombinator.com/item?id=42505496
That being said, the IP blocks are shared by their iGPUs, so the discrete GPUs do not need to recoup the costs of most of the R&D, as it would have been done anyway for the iGPUs.