Senior CPU Performance Engineer

Overview We are looking for a CPU Performance Engineer to join our team, with a focus on understanding CPU performance on real devices and translating observations into insights that can help guide CPU architecture and microarchitecture decisions.

We work with silicon platforms and commercial devices, investigating performance in complex environments where system behaviour is not always predictable or easy to reproduce. This requires a solid understanding of CPU fundamentals, together with an interest in exploring how software and system-level behaviour interact with the underlying CPU microarchitecture.

You will work alongside experienced CPU architects and performance engineers to investigate performance bottlenecks, develop and test hypotheses, and build the evidence needed to support CPU design decisions.

While this is not a CPU design or microarchitecture role, our work directly contributes to their development. We identify and analyse performance limitations observed on devices, validate our findings using pre-silicon platforms (FPGA, emulation or performance models) when possible, and use the results to make concrete, data-backed recommendations to CPU teams.

Responsibilities Analyse Performance on Real-World Workloads We investigate the performance of real-world workloads running on silicon platforms and commercial devices, looking at how system-level behaviour affects CPU performance and identifying the factors that contribute to observed bottlenecks.

Our investigations often take place in complex environments with limited observability, where waveforms are unavailable, performance counters provide only partial information, and measurements can be noisy or difficult to reproduce.

CPU-Focused Performance Analysis We use PMU counters, profiling tools and targeted experiments to understand CPU behaviour and identify performance limitations across different parts of the microarchitecture, including memory, front-end, back-end and execution.

When needed, we extend our investigations to more controlled environments, using FPGA, emulation or performance models to reproduce behaviours observed on silicon, isolate CPU-specific effects and validate hypotheses.

Where applicable, we compare results across silicon, models and emulation platforms to better understand observed differences and build confidence in our conclusions.

Influence CPU Microarchitecture Our findings contribute to discussions with CPU architecture and build teams, providing data and analysis to help find opportunities for improving future CPU compositions.

We document investigations, communicate results and recommendations clearly, and contribute to the continuous improvement of our existing performance analysis tools and workflows.

Experience Required Skills We are looking for someone with:

Good understanding of CPU microarchitecture fundamentals (pipelines, caches, speculation, memory hierarchy).Experience in performance analysis, including profiling, benchmarking or debugging.

Familiarity with PMU-based profiling tools or similar performance counters.

Basic programming and scripting skills (Python, C/C++ or similar).Strong analytical and problem-solving skills.

Nice to have

We would also value familiarity with:

CPU performance analysis, benchmarking and low-level software (C/assembly).RTL concepts, FPGA or emulation platforms.

Performance investigations on real devices (mobile, embedded or client).System-level interactions (OS, scheduling, memory).Arm architecture.

Profile We are looking for someone who:

Is curious about how real-world software and system behaviour affect CPU performance.

Is comfortable investigating complex problems with incomplete or noisy data.

Is eager to learn, collaborate with experienced engineers and develop their technical expertise.

Can communicate findings clearly and support conclusions with evidence.

Additional Information If skilled worker sponsorship is required, Arm will meet costs associated with sponsorship for the employer only. All fees associated with the individual’s application e.g. visa application fee and Immigration Health Surcharge, will be the responsibility of the successful candidate.#LI-CIOur 10x Mindset At Arm, we believe progress happens when people are empowered to think bigger and push beyond what seems possible.

Our 10x mindset is about curiosity, ambition and creating impact that will be used by millions. We learn fast, prioritise collaboration and turn bold ideas into real technology. We look for people who are…