Performance Engineer (Workload Modelling and Simulation)

Huawei’s vision is a fully connected, intelligent world.

To achieve this, we work to inspire passion for basic research around the world.

Our combined passion drives development across the global innovation value chain.

Huawei has the largest Research and Development organization in the world with 96,000+ employees in research centers around the globe.

In the UK, we already have design centers in Cambridge, London, Edinburgh and Ipswich.

We continue to explore and define new research directions and new services.

We have expanded our collaborations with academic researchers; researched new network architectures, integration of communications and key enabling technologies; and developed the fundamental theories of these technologies.

We invite you to join us on this exciting journey and drive your career forward.

We are seeking a highly motivated and enthusiastic Performance Engineer to join our dynamic workload modelling team.

In this role, you will have the opportunity to work on cutting-edge projects involving performance projection, simulation, and architectural studies, with a focus on server CPUs, NPUs, and AI workloads.

As a Performance Engineer, you will contribute to the development of performance models for upcoming server processors and accelerators, support architectural studies, and drive software/hardware co-optimization for next-generation systems.

• Develop and enhance simulation features to enable rapid architectural exploration and performance evaluation of server CPUs and NPUs, focusing on AI and large-scale data analytics workloads.
• Conduct in-depth performance projections for various workloads, including databases, distributed storage, and engines for AI and data analytics.
• Contribute to architectural studies to explore and evaluate the latest server CPU core and SOC designs.
• Work on characterizing workloads and developing methodologies for tracing and optimizing AI models to enhance simulation and performance analysis.
• Construct a non-intrusive, highly accurate system for characterizing and modelling complex workloads, ensuring precise workload representation.
• Collaborate with cross-functional teams to extract and analyze real-world workload features, contributing vital data for hardware development.

• Required:
• Strong understanding of CPU architecture and micro-architecture performance techniques (e.g., branch prediction, prefetchers, cache hierarchies).
• Proficient in performance analysis and workload characterization, with hands-on experience in methodologies for system-level architectural exploration.
• Experience in developing using dynamic binary instrumentation infrastructures like QEMU or DynamoRIO or x86 PIN.
• Proficiency in C/C++, with a solid understanding of Assembly Language.
• Experience with Python and other scripting languages to support automation, data processing, and tool development.
• Excellent analytical and problem-solving skills with the ability to work both independently and as part of a team.
• Desired:
• Experience in compiler technologies, binary analysis, and performance tuning.
• Experience in developing and using performance simulators like GEM5 (O3 model), Sniper or others
• Knowledge of AI workloads and the challenges involved in optimizing large-scale models for performance simulation.
• Experience in Linux kernel development, including knowledge of kernel internals.
• Hands-on experience in CPU performance analysis, utilizing methodologies such as PMU-based profiling and TopDown Analysis, and proficiency with performance analysis tools like Linux perf.

What we offer

• 33 days annual leave entitlement per year (including UK public holidays)

• Life insurance
• Private medical insurance
• Medical expense claim scheme

• Cycle to work scheme
• Company sports club and social events
• Additional time off for learning and development