CUDA 13.4 Explained: New Tile Features, PTX Instructions and Rubin sm_107 Preview
NVIDIA Developer · youtube · 2026-10-03
NVIDIA's CUDA team walks through CUDA 13.4: new CUDA Tile features and programmatic dependent launch (PDL) usable from cuTile Python and Tile IR; new PTX instructions including L1 prefetch, int4-to-int8 matrix loads, packed FP8 arithmetic and lighter proxy fences; a developer preview targeting Rubin's sm107 so code can be ported before launch, plus locality-aware APIs for multi-die GPUs; and cgroup-based GPU memory limits with an updated MPS for safer containerized GPU sharing, alongside cuBLAS, NVCC and tooling updates.
More from Infra
- Suhail hails the start of the Vera Rubin era as NVIDIA's next-gen GPU boots up — rickasaurus · 2026-10-03
- GPU-backed loans put GPU earning power and collateral value under scrutiny — AnneliesGamble · 2026-10-03
- Google accused of illegally bulldozing 300 million sq. meters of Finnish forest for AI data centers — Polymarket · 2026-10-03
- SkyRL v0.4 trains 1T-param Kimi K2.7 with RL on just 16 B300 GPUs — casper_hansen_ · 2026-10-03
- BIS report: 55% of AI investment is circular, echoing Lucent-Nortel era risks — rohanpaul_ai · 2026-10-03
- RAM prices may 10-20x next year and double again by 2028, predicts analyst — Kyrannio · 2026-10-03