GPU Compiler Engineer

Morr0; San Francisco (CA); On-site; USD 270,000 - 330,000; Full time

Job summary

Morr0 is seeking a compiler engineer for Campbell, CA on-site to build an LLVM/MLIR-based compiler for a new GPU architecture. You will work with hardware architects and GPU software engineers to optimize performance across the stack.

Required are strong C++ skills, hands-on compiler experience, and familiarity with LLVM/MLIR ecosystems. The role offers growth with competitive base pay plus equity in a cutting-edge startup.

Qualifications

• Strong programming in C++.
• Hands-on experience developing compiler infrastructure or related low-level software.
• Familiarity with LLVM, MLIR, Triton, TVM, IREE or similar compiler technologies.
• Understanding of compiler concepts such as IRs, optimization passes, code generation.
• Interest in GPU architectures or hardware-aware software development.
• Experience with GPU compiler backends and performance optimization is a plus.

Responsibilities

• Developing compiler infrastructure using LLVM, MLIR or similar frameworks.
• Implementing compiler passes, IR transformations, lowering pipelines and code generation.
• Optimizing GPU execution through scheduling, register allocation, memory management and instruction-level improvements.
• Supporting GPU compute workloads, kernel compilation and hardware-specific optimizations.
• Collaborating directly with hardware architects and GPU software engineers to improve performance.
• Debugging, profiling and benchmarking compiler-generated code.

Skills

C++

Compiler infra

LLVM/MLIR familiarity

GPU backends experience

Performance optimization

Tools

LLVM

MLIR

Triton

TVM

IREE

CUDA

Job description

Campbell, CA | On-site | Full-time | Junior/Mid/Senior Seniority Levels

Build the compiler behind a new GPU architecture.

An ambitious semiconductor startup developing a new GPU architecture and the software stack to power it. Founded by a world-renowned GPU architect, the company is bringing together an exceptional team of hardware architects, compiler engineers and GPU software specialists.

This is an opportunity to work on an LLVM/MLIR-based compiler targeting custom GPU hardware, with direct exposure to the engineers designing the architecture itself.

What you'll be working on

• Developing compiler infrastructure using LLVM, MLIR or similar frameworks.
• Implementing compiler passes, IR transformations, lowering pipelines and code generation.
• Optimizing GPU execution through scheduling, register allocation, memory management and instruction-level improvements.
• Supporting GPU compute workloads, kernel compilation and hardware-specific optimizations.
• Collaborating directly with hardware architects and GPU software engineers to improve performance.
• Debugging, profiling and benchmarking compiler-generated code.

What we're looking for

• Strong programming skills in C++.
• Hands‑on experience developing compiler infrastructure or related low-level software.
• Familiarity with LLVM, MLIR, Triton, TVM, IREE or similar compiler technologies.
• An understanding of compiler concepts such as intermediate representations, optimization passes, code generation or instruction scheduling.
• An interest in GPU architectures, AI accelerators or hardware‑aware software development.

Experience with GPU compiler backends, CUDA, graphics/shader compilers, custom accelerators or performance optimization would be particularly relevant.

You don't need to tick every box. We're open to engineers at different stages of their careers, including those with just a couple of years of relevant experience who are keen to develop their expertise.

Base salary: Up to $300k base salary + equity and benefits.

Recruitment managed by Morr0 Labs on behalf of a semiconductor client.