(Senior) Researcher - Computer Architecture ML Systems

Huawei Switzerland; Zürich; Vor Ort; CHF 90’000 - 130’000; Vollzeit

Zusammenfassung

A leading global technology firm is seeking Researchers for its Computer Architecture Innovation Lab in Zurich. The role involves developing advanced NPU-centric frameworks, collaborating with hardware teams, and driving impactful research in the field. Ideal candidates will have a PhD or MSc in computer science and proficiency in AI frameworks and programming languages.

Leistungen

Competitive salary and incentive schemes

Research on high-impact topics

Work with top Researchers and University Professors

International mobility

Qualifikationen

• PhD or MSc in computer science or related field.
• 3+ years in system software development.
• Experience with compiler frameworks and NPU architectures is a plus.

Aufgaben

• Architect and implement Python frameworks for AI workloads targeting NPU.
• Design NPU-specific compiler stacks to optimize performance.
• Collaborate with hardware teams on microarchitecture features.

Kenntnisse

Python

C/C++

AI frameworks

NPU architectures

Ausbildung

PhD or MSc in Computer Science

Tools

LLVM

GCC

TVM

TensorFlow

PyTorch

Jobbeschreibung

Huawei is a leading global information and communications technology (ICT) solutions provider. Driven by a commitment to sound operations, ongoing innovation, and open collaboration, we have established a competitive ICT portfolio of end-to-end solutions in telecom and enterprise networks, devices, and cloud technology and services. Our ICT solutions, products, and services are used in more than 170 countries and regions, serving over one-third of the world's population. With 180,000 employees, Huawei is committed to enabling the future information society, and building a Better Connected World.

Huawei's Switzerland Research Centre in Zurich is responsible for advanced technical research, architecture evolution design and strategic technical planning on computer architecture.

We are currently looking for Researchers of our new Computer Architecture Innovation Lab.

Job Responsibilities:

• NPU-Centric Python Framework Design：Architect and implement Python frameworks (e.g., PyTorch/TensorFlow integrations) to optimize NPU utilization for AI workloads. Develop tools for automated model conversion, quantization, and deployment targeting NPU subsystems.
• Compiler & Toolchain Innovation：Design and enhance NPU-specific compiler stacks (e.g., LLVM/TVM/GCC-based) to translate high-level models into optimized NPU instructions; Optimize kernel scheduling, memory allocation, and parallelism strategies for NPU architectures.
• Hardware-Software Co-Design: Collaborate with hardware teams to define NPU ISA extensions, microarchitecture features, and performance counters for framework efficiency. Performance Analysis & Optimization
• Profile end-to-end AI workflows (training/inference) to identify bottlenecks in NPU-Python frameworks. Implement low-latency, high-throughput solutions for transformer-based models and generative AI workloads.
• Research & Ecosystem Leadership: Publish cutting-edge research in top-tier conferences (e.g., ISCA, ASPLOS, MLSys) and contribute to open-source projects; Drive adoption of NPU frameworks through developer tools, documentation, and industry partnerships.

Requirements and qualifications:

• PhD or MSc in computer science or area related to computer architecture, or Compiler Engineering experience in industry
• Proficiency in Python and C/C++ for system-level programming.
• Proficiency in AI frameworks (PyTorch, TensorFlow) and model optimization techniques.
• 3+ years in system software development,
• Experience with xPU/DSP toolchains is a plus.
• Knowledge of compiler frameworks (LLVM, GCC, TVM, XLA) and NPU/GPU architectures is a plus.
• Excellent oral and written English.

What we offer

• Competitive salary and incentive schemes
• Research on high-impact topics
• Work with top Researchers and University Professors
• International mobility

Application Process:

To apply, please click on the link