Anthropic's Claude Mythos generates AMD GPU kernels autonomously, reducing 12-month workload to 3 days Anthropic's Claude Mythos model autonomously wrote nearly 100 percent of AMD GPU compute kernels, cutting software onboarding work from up to 12 months to roughly 3 days, according to Tom's Hardware. The result points to autonomous code generation for GPU optimization and reduces reliance on NVIDIA's CUDA ecosystem. Anthropic's Claude Mythos generates AMD GPU kernels autonomously, reducing 12-month workload to 3 days According to Tom's Hardware, Anthropic's Claude Mythos model autonomously generated nearly 100 percent of AMD GPU compute kernels, reducing software onboarding work from up to 12 months to approximately 3 days. The achievement demonstrates autonomous code generation capabilities applicable to GPU optimization work and reduces reliance on NVIDIA's CUDA ecosystem. Topics Sources - Press Read article https://www.tomshardware.com/tech-industry/artificial-intelligence/anthropic-says-claude-mythos-slashed-software-work-to-onboard-amd-gpus-from-up-to-12-months-to-around-three-days-writing-almost-100-percent-of-kernels-autonomously-chips-away-at-nvidias-cuda-moat Go deeper This intelligence is sourced automatically from public sources across the web and synthesised by the Prefactor AI pipeline. Stories are reviewed before publication.