Deep Learning Compiler Engineer - CUDA
Core
Design and implement the DSL and core compiler for a tile-aware GPU programming model to optimize performance on emerging GPU architectures.
Role type
Senior IC compiler architect (GPU/CUDA)
Builds
Next-generation GPU architectures, DSLs, and compiler stacks for AI/LLM workloads
Domain
Hardware/Software co-design, GPU architecture, High-Performance Computing (HPC)
Deliverable
production ML models | infrastructure
Required skills
C/C++ programming, computer architecture fundamentals, compiler development (MLIR/TVM/Triton/LLVM), GPU kernel programming
Preferred skills
LLM algorithms, multi-GPU distributed communication, abstract problem solving
Technologies
CUDA, MLIR, TVM, Triton, LLVM
Responsibilities
Design and implement DSLs and core compilers for tile-aware GPU programming models; Innovate and iterate on compiler core architecture to optimize performance; Investigate next-generation GPU architectures and provide solutions in the DSL and compiler stack; Perform performance analysis on emerging AI/LLM workloads and integrate with AI/ML frameworks
Seniority
Senior, hands-on IC
