AMD is actively marketing the Radeon RX 7000 range of graphics cards for artificial intelligence-related jobs. The flagship model RX 7900 XTX showed impressive results and surpassed the RTX 4090 in some areas in tests using DeepSeek AI.
AMD's analysis shows that the RX 7900 XTX performs 34% better than the RTX 4080 SUPER in DeepSeek-R1-Distill-Qwen-7B, 27% better in DeepSeek-R1-Distill-Llama-8B, and 22% better in DeepSeek-R1-Distill-Qwen-14B. Although less pronounced, the difference with the RTX 4090 is nevertheless discernible: in certain tests, the RX 7900 XTX exhibits performance advantages ranging from 2% to 13%.
The RX 7900 XTX is a fantastic option for working with neural networks and machine learning because it has 24 GB of video memory. By providing guidance on configuring Ryzen AI and RX 7000 series graphics cards to function with DeepSeek R1, AMD actively supports such tasks.
Crucially, AMD's less expensive models—the RX 7600 XT, RX 7700 XT, and RX 7800 XT—can manage AI models that aren't as demanding, enabling users who work with machine learning. When it comes to high-level AI solutions, NVIDIA continues to lead in performance, while AMD is making great progress with strong and reasonably priced neural networking substitutes.
Main image: reddit.com
0 comments