Senior AI Compute Architect — GPUs, MLIR & CUDA

Oho Group; San Francisco (CA); On-site; USD 260,000 - 380,000; Full time

Job summary

Oho Group is hiring a senior architect to shape the software stack for a programmable accelerator platform. You will connect models, frameworks, compilers, runtimes and hardware from high-level workload behavior to kernel execution and code generation.

You will work across graph capture, MLIR/LLVM lowering, runtime and driver architecture, with a focus on SIMT execution, memory hierarchy and performance profiling, guiding senior engineers through design reviews and pre-silicon validation.

Qualifications

• 12+ years across AI, compiler, runtime or systems software.
• Strong C++ with LLVM and MLIR experience.
• Expertise in GPU/SIMT execution and CUDA programming.
• Understanding of modern AI models at graph and operator level.
• Experience connecting frameworks, compilers, runtimes and hardware.
• Hands-on, measurement-led approach to architecture.
• Experience guiding senior engineers through design and code reviews.

Skills

C++

LLVM

MLIR

GPU/SIMT execution

CUDA programming

Framework integration

Performance profiling

Tools

PyTorch

XLA

Triton

CUDA

MLIR

Job description

Oho Group is hiring a senior architect to shape the software stack for a programmable accelerator platform. You will connect models, frameworks, compilers, runtimes and hardware from high-level workload behavior to kernel execution and code generation.

You will work across graph capture, MLIR/LLVM lowering, runtime and driver architecture, with a focus on SIMT execution, memory hierarchy and performance profiling, guiding senior engineers through design reviews and pre-silicon validation.