Compiler Performance Engineer
Core
Analyze and optimize compiler-generated code for Intel CPUs and GPUs, tune workloads, and implement compiler components to maximize hardware performance for AI, HPC, and demanding workloads.
Role type
Compiler Performance Engineer
Builds
High-performance, portable software for heterogeneous computing (CPUs, GPUs, accelerators)
Domain
Semiconductor / Compiler Technology / HPC / AI
Deliverable
production ML models
Required skills
C/C++/Assembly/Python, compiler theory (parsing, IR, data-flow analysis, code generation), performance analysis, test automation, open-source contribution
Preferred skills
Intel x86/AVX or GPU architecture knowledge, LLVM/MLIR/GCC, parallel/heterogeneous programming (OpenMP, SYCL, CUDA), AI/ML frameworks
Technologies
LLVM, MLIR, SYCL, Triton, CUDA, OpenMP, OpenCL, PyTorch
Responsibilities
Analyze performance issues in compiler-generated code and root-cause optimization phases; Tune workloads and measure performance on Intel CPUs and GPUs; Design and implement test automation for continuous integration; Connect hardware/compiler capabilities with AI/HPC software ecosystems; Participate in upstream open-source work; Implement and extend compiler components (optimization passes, code generation, runtime, offload support)
Seniority
Experienced Hire
