AIPhoronixabout 1 hour ago
AMD Tests Porting 50 LLMs To Client GPUs
1 min read
AMD engineers led an experiment to port 50 popular Hugging Face models across 18 architectures to AMD client GPUs in just three to four days. The team found a 1.5 to 2.1 times speed-up for 32K token prompt processing compared to Llama.cpp's ROCm back-end.
Level
Hype check
Tap to vote and see what everyone thinks.
In this storyAMD
