Senior Software Engineer - GPU Kernel Authoring & Optimization

coreweave · Sunnyvale, CA / Bellevue, WA

Posted
1 day ago
Last confirmed live
Today

What this role involves

CoreWeave is seeking a Senior Software Engineer for GPU Kernel Authoring & Optimization to write, profile, and tune CUDA kernels for large-scale model serving. The role involves optimizing for hardware, benchmarking, and contributing to MLPerf runs. The engineer will lead designs and partner with cross-functional teams to improve latency, throughput, and reliability.

Skills this posting asks for

  • cuda
  • gpu kernel authoring
  • optimization
  • gemm
  • attention
  • moe routing
  • quantization
  • kv-cache
  • fused epilogues
  • tensor cores
  • memory coalescing
  • shared-memory
  • register usage
  • kernel-authoring dsls
  • compilers
  • microbenchmarks
  • roofline analyses
  • vllm
  • tensorrt-llm
  • llm-d
  • sglang
  • mlperf
  • inference
  • training

Requirements

  • Level: senior

From the employer’s posting

CoreWeave is The Essential Cloud for AI™. Built for pioneers by pioneers, CoreWeave delivers a platform of…

Read the full description on coreweave’s careers page

Apply without filling the form

Approve this role and the application is completed for you, including a résumé tailored to it. You get a confirmation when it lands, and a credit is only spent when a submission is confirmed.

Other roles at coreweave

All 88 roles at coreweave