Member of Technical Staff, GPU Compiler
$285k - $315k • San Francisco • FullTime
Posted 1mo ago
About the job
We are seeking a Member of Technical Staff for GPU Compiler Engineering to build the machine that drives our cutting-edge GPU compiler. This role involves working on the entire pipeline, from StableHLO ingestion to backend code generation and executable binary creation. You will enhance our search-based compiler infrastructure, enabling agents and RL to find optimal code forms while guaranteeing correctness. This is an opportunity to work at the foundational level of the AI stack, directly impacting the speed and cost of large-scale training runs.
Responsibilities
- Build and extend MLIR dialects and passes for optimizing training and inference workloads.
- Work within the LLVM backend to optimize instruction selection, scheduling, register allocation, and direct cubin emission.
- Develop equivalent optimization depth for AMD, TPU, and Trainium targets.
- Expand search-based compiler infrastructure, including agent- and RL-driven program search and formal correctness proofs.
- Implement classic compiler optimizations tailored for large-scale training.
- Create hybrid codegen paths for complex lowering scenarios.
- Own testing, benchmarking, and performance regression systems, including hardware model implementation.
- Collaborate with the research team and customers to identify critical optimization opportunities.
Requirements
- Deep experience in compiler infrastructure (LLVM, MLIR, or similar).
- Strong background in GPU architecture and low-level optimization (CUDA, ROCm, or similar).
- Hands-on experience with PTX/SASS, GCN/RDNA assembly, or other GPU ISAs.
- Familiarity with ML compiler stacks (XLA, TVM, Triton, torch.compiler, or similar).
- Solid systems programming skills in C++ and/or Rust.
- Proven track record of building production-grade compiler infrastructure.
Benefits
- Meaningful equity
- Relocation assistance