GPU Compiler Engineer – MLIR/LLVM Backends

OXMIQ Labs, Inc.; Campbell (CA); On-site; USD 140,000 - 210,000; Full time

Benefits offered by this job

Equity participation

Medical, dental, and vision coverage

Job summary

OXMIQ Labs, Inc. is seeking a senior compiler engineer to design and implement GPU-focused compiler infrastructure, including MLIR-based passes and LLVM backends for Oxmiq hardware and GPUs.

You will optimize code generation, work with architecture teams on performance strategies, and collaborate with kernel engineers on Triton-based models and custom op lowering. Strong C++/LLVM, MLIR, and CUDA experience are required.

Qualifications

• 5+ years of compiler engineering experience focused on GPU or hardware backends.
• Strong C++ and LLVM expertise, with LLVM IR and pass infrastructure familiarity.
• Hands-on MLIR experience including dialect design and lowering.
• Proven GPU programming experience (CUDA) and performance optimization knowledge.

Responsibilities

• Design and implement compiler infrastructure for GPU workloads.
• Develop and optimize MLIR-based passes for Oxmiq hardware IP and GPUs.
• Build and maintain LLVM backend components (instruction selection, register allocation).
• Collaborate with architecture teams to translate performance requirements into codegen strategies.
• Optimize compiler output for throughput, latency, and memory efficiency.
• Collaborate with kernel engineers to support Triton-based models and custom op lowering.
• Participate in design reviews, code reviews, and architecture discussions.
• Support performance profiling, benchmarking, root cause analyses for compiler code.

Skills

C++

LLVM

MLIR

GPU backends

CUDA

Code generation

Performance profiling

Debugging

Claude Code

Education

BS/MS/PhD in Computer Science/Engineering

Tools

MLIR

Triton

CUDA toolkit

Python

Job description

OXMIQ Labs, Inc. is seeking a senior compiler engineer to design and implement GPU-focused compiler infrastructure, including MLIR-based passes and LLVM backends for Oxmiq hardware and GPUs.

You will optimize code generation, work with architecture teams on performance strategies, and collaborate with kernel engineers on Triton-based models and custom op lowering. Strong C++/LLVM, MLIR, and CUDA experience are required.