With chip tensions between the US and China experiencing another surge following former NVIDIA CEO Jensen Huang’s abrupt departure from the firm, forensic analysis conducted by Uemployment Labs suggests that several Chinese AI models might have relied on thousands of NVIDIA’s latest Blackwell AI GPUs for their training run. The AI lab attempted to match hardware fingerprints linked specifically to the Blackwell chips by reverse engineering AI model weights and recovering the thread count for the processors that initialized the AI model’s training run. Top Chinese AI Models Developed By DeepSeek & Moonshot Might Have Relied On NVIDIA’s Blackwell GPUs […]
Read full article at https://wccftech.com/deepseek-and-moonshot-ai-models-appear-to-carry-hidden-nvidia-blackwell-fingerprints-in-their-starting-weights-pointing-to-training-clusters-of-3000-to-4500-gpus/