Nvidia has launched Nemotron 3 Ultra, a large language model optimized for high performance on the RTX platform. According to evaluations from Artificial Analysis, it is currently the most capable open-source AI model from the US, surpassing domestic rivals in both speed and accuracy.
Context
The release of Nemotron 3 Ultra is part of Nvidia's strategy to provide powerful AI solutions that can run locally or on enterprise cloud infrastructure. This model is specifically trained to maximize the computational capabilities of Nvidia GPUs, significantly reducing operational costs and response times for real-world applications.
Why it matters
Despite leading in the US, Nemotron 3 Ultra still faces stiff competition from Chinese open-source models like Qwen2 or DeepSeek, which continue to dominate many global benchmark rankings. This highlights the intense open-source AI race between the two tech superpowers, with Nvidia being the heaviest contender from the US side.