Recent news highlights NVIDIA's RTX 5090 in the DeepSeek R inference performance benchmarks, where it has demonstrated exceptional prowess. The RTX 5090 boasts at least double the performance of AMD's RX 7900 XTX. Across multiple tests with DeepSeek R1 models, the RTX 5090 not only significantly outperformed the RX 7900 XTX but also showed marked improvement over its predecessor, the RTX 4090. Notably, in the Distill Qwen 7b and Distill Llama 8b models, the RTX 5090 processes 200 Tokens per second, nearly twice as fast as the RX 7900 XTX.
