AI Compiler Optimization Engineer

About The Company

Huawei Research and Development UK Limited is part of Huawei, a global leader in information and communications technology (ICT) infrastructure and smart devices, established in 1987. With over 207,000 employees worldwide, Huawei operates in more than 170 countries, serving upwards of three billion people. The company is driven by a vision to connect every person, home, and organization with intelligent digital solutions, fostering ubiquitous network access, AI-enhanced computing power, and transformative digital platforms.

Huawei UK hosts research centers in Cambridge, London, Edinburgh, and Ipswich, emphasizing close collaboration with leading academic institutions to pioneer advancements in network architectures and communication technologies.

Role Overview

We are recruiting an AI Compiler Optimization Engineer specialized in enhancing AI model inference efficiency by leveraging advanced compiler technologies. The role focuses on optimizing performance for CPU and hybrid CPU/XPU architectures by profiling AI frameworks to detect improvement opportunities and contributing insights from the latest industry research.

Key Responsibilities

• Develop and apply compiler optimization techniques, including MLIR and LLVM backend enhancements, to improve inference speed on CPU and CPU/XPU systems.
• Refine just-in-time (JIT) computation graphs through operator fusion, efficient memory management, and other methods to boost latency and throughput.
• Analyze full inference workflows within AI frameworks such as TensorFlow, PyTorch, ONNX, and llama.cpp to identify bottlenecks and performance hotspots.
• Formulate and implement tuning strategies including kernel adjustments and graph-level improvements to elevate performance.
• Continuously monitor state-of-the-art AI and compiler research from academic publications and open-source projects to derive actionable insights and innovation.
• Generate comprehensive technical reports summarizing trends, benchmarks, and optimization opportunities, requiring strong technical writing aptitude.
• Perform additional tasks as needed under managerial guidance, with job scope evolving alongside Huawei UK's strategic needs.

Candidate Profile

Essential Qualifications & Skills:

• Proficiency in C/C++ programming and experience with compiler infrastructures like LLVM, MLIR, or comparable platforms.
• Thorough understanding of AI model structures and inference execution.
• Hands-on experience with profiling and performance tools such as perf, TensorBoard, and VTune.
• Knowledge of CPU and XPU hardware designs and relevant optimization methodologies.
• Excellent analytical and problem-solving abilities.

Preferred Attributes:

• A bachelor's or master's degree in Computer Science or related disciplines.
• Participation in open-source compiler initiatives, notably within LLVM or MLIR communities.
• Experience working with heterogeneous computing environments encompassing CPU, GPU, and XPU.
• Publication of technical articles, papers, or blogs pertaining to AI or compiler optimizations.
• Self-motivated learner capable of independent exploration and adept practical skills.

Employee Benefits

• 33 days of annual leave including UK public holidays.
• Group personal pension scheme.
• Life and private medical insurance coverage.
• Medical expense claim program.
• Employee Assistance Program for wellbeing support.
• Cycle-to-work scheme promoting sustainable commuting.
• Access to company sports clubs and social activities.
• Additional leave allocated for learning and development pursuits.