- AMD reference points show RX 7900 XTX Beating RTX 4090, 4080 Super in AI tests
- Nvidia replied, claiming that RTX 5090 is 2.2x faster than AMD GPU
- The reference points differ, but RX 7900 XTX of AMD is much cheaper than Nvidia cards
Depseek is the new AI Darling, at least for now, and Nvidia and AMD have been discussing which of them he directs it faster.
AMD began the dispute by launching reference points that show its high -end graphics card RX 7900 XTX that exceeds RTX 4090 and RTX 4080 Super in the R1 Deepseek R1 tests. Nvidia inevitably responded with his own results that show the opposite. Because, of course, they do.
According to David Mcafee in X, the AMD -based GPUs was up to 13% faster than RTX 4090 and 34% ahead of the RTX 4080 Super. The RX 7900 XTX worked better against RTX 4090 using Deepseek R1 Distill Qwen 7B, where it led by 13%. AMD also tried distill calls 8b and distill Qwen 14b, with its GPU 11% and 2% faster, respectively. The RTX 4090 had an advantage, it was 4% ahead in distill qwen 32b.
Against the RTX 4080 Super, the AMD GPU showed larger cables. The RX 7900 XTX was 34% faster using Deepseek R1 distill qwen 7b, 27% ahead with distill calls 8b and 22% faster using distill qwen 14b. Obviously, we cannot be sure how the NVIDIA GPUs were configured for the tests and, as AMD executed them, it is fair to say that Team Red may not have done everything possible to optimize things for the hardware of Team Green.
Nvidia defends himself
As Tom hardware He points out, RX 7900 XTX is not widely used for AI, but its RDNA3 architecture includes AI and AMD processing capabilities has commercialized this aspect under the “AI Accelerator” label.
Shortly after AMD reference points were made live, Nvidia responded in a blog post with their own results, also reported by Tom hardwareclaiming the RTX 5090 is up to 2.2 times faster than RX 7900 XTX. Using Qwen 32b, Nvidia reported a 124% advantage, while RTX 4090 was 47% ahead. With flame 8b, RTX 5090 was 106% faster, and RTX 4090 led by 47%.
The round trip inevitably highlights the importance of treating the reference points released to the manufacturer with caution. Different optimizations, driver versions and test conditions can lead to different results.
It is also important to keep in mind that even if Nvidia’s reference points are true, AMD still comes out at top in a very important aspect: the price. Its GPU is significantly cheaper than both Nvidia offers.
