attention head AI jobs Alerts Saved
AI jobs › AI jobs using Triton

AI jobs using Triton

Triton, from OpenAI, lets engineers write GPU kernels in Python-like code and compiles them to run efficiently, taking care of much of the tiling and memory work CUDA leaves to the programmer. PyTorch's compiler generates Triton kernels, and many fast attention and quantization kernels are written in it.

Roles naming it are kernel and performance engineering, speeding up training and inference, usually beside CUDA. NVIDIA's Triton Inference Server, a model server, shares the name, and some postings that name Triton mean that instead.

398 open AI roles at 219 employers, from their own job boards. Showing the newest 100.

NVIDIA · 51Amazon · 15Cerebras · 8Inferact · 5Nebius · 5Pragmatike · 5Centific Global Solutions, Inc. · 4CoreWeave · 4DigitalOcean · 4eBay Inc. · 4Fundamental · 4KRAFTON · 4Microsoft · 4Rhoda AI · 4Together AI · 4

Related