Performance Modeling Architect

Position: Performance Modeling Architect

Location: Cambridge, UK (100% Onsite)

Employment Type: Full-time

Overview:

Join a cutting-edge CPU Architecture team leading system software innovation for next-gen Cloud and Mobile solutions. As a Performance Modeling Architect, you’ll work closely with CPU and system architects to bring architectural features to life through firmware, kernel, hypervisor, and application integration. This is a high-impact role blending hands-on system software development with strategic technical leadership.

Key Responsibilities:

• Serve as the technical lead across the full software stack (firmware, kernel, hypervisor, drivers, applications).
• Design, prototype, and deliver system-level software solutions, balancing performance, quality, and security.
• Drive low-level system programming and OS design in C/C++.
• Integrate architectural features into OS-level components and drivers.
• Collaborate with open-source communities to drive adoption of new features.
• Support R&D, performance optimization, and real-world deployment of Arm-based CPU advancements.
• Lead profiling, instrumentation, and performance tuning across diverse workloads.

Required Skills & Experience:

• 5+ years in system software development (kernel, hypervisor, low-level programming).
• Expertise in C/C++/assembly and Python.
• Strong experience with Linux kernel, drivers, OS internals, and firmware.
• Deep understanding of Arm CPU architecture and SoC systems.
• Proven leadership in system software projects with hands-on code contributions.
• Knowledge of upstream Linux collaboration and open-source integration.
• Familiar with profiling tools, optimization strategies, and system performance measurement.
• Excellent communication skills across cross-functional teams.

Preferred Qualifications:

• M.S. or Ph.D. in Computer Science, Electrical Engineering, or related.
• In-depth knowledge of AArch64 architecture and virtualization.
• Experience with Android or HarmonyOS.
• Exposure to heterogeneous computing (GPU, NPU, ISP, compositing).