OpenAI releases Triton
Released Triton — researchers with no CUDA experience can write cuBLAS-level FP16 matmul kernels in about 25 lines
Why it matters
Opened kernel writing up from a handful of experts to ordinary researchers
Source — Primary sources first; secondary reports are labelled “Press”
Officialopenai.comhttps://openai.com/index/triton/Added in Week 35, 2026 · See that week