PyTorch 2.14 ships NVGEMM CUTLASS kernels, nccl2 backend and c10d fault tolerance
PyTorch · x · 2026-09-03
PyTorch 2.14 is out. Highlights:
- NVGEMM: CuTeDSL-generated CUTLASS kernels in Inductor, with epilogue fusion, scaled/NVFP4 GEMM, grouped-reduction epilogues autotuned alongside Triton and ATen
- Distributed: new nccl2 backend ported from torchcomms, full collective contract with nonblocking communicators and eager communicator splitting
- Fault tolerance is now a first-class c10d concept with in-place process-group reconfiguration and one-sided RMA windows
- Also: native linear algebra on Apple Silicon, torch.switch, CUDA graph capture for torch.whileloop, declarative dynamic shapes via @dynamicspec, and experimental torch.compile for complex-valued tensors
Related event: PyTorch 2.14 Released with 2,995 Commits from 487 Contributors(3 posts)→
More from coding & agent
- Anthropic borrows lawyers' ethics for Claude agents that closed 186 deals with no human approval — rohanpaul_ai · 2026-10-18
- One prompt turns Claude Code into a full motion-graphics video pipeline with Remotion — aziz4ai · 2026-10-18
- Prompt injection via email can hijack your AI agent — jev-guard adds risk-scored guards — sujingshen · 2026-10-18
- 14-Agent AI Film Pipeline on One RTX 5090: Fixing Cut Defects Between Clips — developervkmp · 2026-10-18
- Meta paper introduces "agent plasticity": best-performing agent isn't the most efficient learner — rohanpaul_ai · 2026-10-18
- Dev ships a bushfire prediction platform in a weekend with frontier models — CtrlAltDwayne · 2026-10-18