AI Compiler Optimization Engineer (Hybrid CPU/XPU)

Huawei Technologies Research & Development (UK) Ltd; City of Edinburgh; On-site; GBP 85,000 - 110,000; Full time

Benefits offered by this job

33 days annual leave entitlement peryr

Group Pension

Life insurance

Private medical insurance

Medical expense claim scheme

Employee Assistance Program

Cycle to work scheme

Company sports club and social events

Extra learning & development time

Job summary

Huawei Research and Development UK Limited in Edinburgh is seeking a skilled AI Compiler Optimization Engineer to boost AI model inference performance on CPU and CPU/XPU hybrid systems, using advanced compiler techniques.

You will profile frameworks such as TensorFlow, PyTorch and ONNX, optimize graph execution, and contribute to open research with practical insights and publications.

Qualifications

• Proficiency in C/C++ and compiler infrastructure (LLVM/MLIR).
• Deep understanding of AI model architectures and inference workflows.
• Experience with performance profiling tools (e.g., perf, TensorBoard, VTune).
• Familiarity with CPU/XPU hardware architectures and optimization techniques.

Responsibilities

• Compiler-Based Performance Optimization: Implement compiler techniques (e.g., MLIR level optimizations, LLVM backend optimizations) to enhance inference performance on CPU and CPU/XPU hybrid systems.
• AI Model Profiling & Framework Optimization: Profile end-to-end inference workflows on frameworks like TensorFlow, PyTorch, ONNX, and llama.cpp to identify hotspots and bottlenecks.
• Research & Insight Development: Track and analyze the latest advancements in AI & compiler research; produce actionable insight reports summarizing trends and benchmarks.

Skills

C/C++

AI architectures

Performance profiling

Education

BSc/MSc/MSci in CS

Tools

LLVM

MLIR

Perf

TensorBoard

VTune

Job description

Huawei Research and Development UK Limited in Edinburgh is seeking a skilled AI Compiler Optimization Engineer to boost AI model inference performance on CPU and CPU/XPU hybrid systems, using advanced compiler techniques.

You will profile frameworks such as TensorFlow, PyTorch and ONNX, optimize graph execution, and contribute to open research with practical insights and publications.