Hosted on MSN17d
DeepSeek's AI breakthrough bypasses Nvidia's industry-standard CUDA, uses assembly-like PTX programming insteadD eepSeek made quite a splash in the AI industry by training its Mixture-of-Experts (MoE) language model with 671 billion parameters using a cluster featuring 2,048 Nvidia H800 GP ...
DeepSeek is generating unprecedented momentum for China's efforts to reduce its reliance on the ecosystem of Nvidia, as ...
DeepSeek's entry into the AI market triggered a ripple effect, with a potential major impact on chipmakers. Shares of Nvidia, ...
DeepSeeks artificial intelligence (AI) models are giving Chinese chipmakers like Huawei a competitive edge against U.S.
With demand for AI accelerators still booming and AMD's market share small, there should be plenty of room for the company to grow its AI business quickly. However, 2025 could be a less-than-stellar ...
These modifications go far beyond standard CUDA-level development ... that the need for high-performance hardware for new AI models would get lower, hurting the sales of companies like Nvidia.
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback