Back to All Open Roles
Systems Engineering
Principal GPU Kernel Engineer
San Francisco, CA / RemoteFull-Time
We are seeking a senior practitioner to join our core systems team. In this role, you will design, profile, and optimize high-throughput distributed tensor execution pipelines.
Key Responsibilities:
- Optimize CUDA / Triton / C++ GPU kernel execution for LLM KV cache memory quantization.
- Architect zero-copy RDMA transport protocol layers across multi-region GPU clusters.
- Collaborate with security researchers to implement confidential compute enclaves.
Role: Principal GPU Kernel EngineerStep 1 of 3