GPU Compute & MLIR Compiler Engineer — AI Workloads
Core
Design, optimize, and deploy high-performance GPU compute kernels and MLIR compiler backends for AI/ML workloads.
Role type
Senior IC GPU compute & MLIR compiler engineer
Builds
High-throughput AI/ML workloads on modern GPU architectures
Domain
Systems Engineering / GPU Compute / Compiler Development
Deliverable
production ML models
Required skills
MLIR framework expertise, OpenCL programming, Vulkan compute programming, C/C++, GPU architecture knowledge, kernel profiling, machine learning fundamentals
Preferred skills
IREE experience, open-source MLIR/IREE contributions, quantization techniques, multi-target compilation
Technologies
MLIR, OpenCL, Vulkan, C/C++, Linalg, Tensor, Tosa, StableHLO, ONNX-MLIR, LLVM, SCF, MemRef, GPU, IREE
Responsibilities
Develop and optimize GPU compute kernels for OpenCL and Vulkan; Design and extend MLIR dialects across abstraction levels; Implement compiler passes and lowering pipelines; Conduct profiling and bottleneck analysis; Build GPU runtime infrastructure; Collaborate on end-to-end model lowering for CV and LLM workloads.
Seniority
Senior, hands-on IC
