GPU Performance Engineer: Optimize & Benchmark Next-Gen SoC

Lex; Greater London; On-site; GBP 90,000 - 120,000; Full time

Job summary

Apple in the United Kingdom seeks a GPU Performance Analysis Engineer to help design and optimize next‑generation GPUs across Apple products. You will work with architecture, software, and hardware teams to diagnose bottlenecks and propose optimisations, developing study methods and measurement tools.

Responsibilities include validating GPU performance through tests, analysing real content, and quantifying benefits of hardware/software features.

Qualifications

• Architecture knowledge of main SoC components (GPU/CPU/interconnect/memory hierarchy).
• Strong hardware and software debugging skills.
• Strong data analysis skills relating performance data to hardware behavior.
• Proficiency in Python and/or C++.
• Ability to plan tasks, estimate deadlines and work in a multi-site team.
• Excellent communication, self-motivation and organisation.
• Knowledge of hardware design.

Responsibilities

• Developing performance test plans to validate GPU performance.
• Analyzing performance issues using micro benches and real content to identify bottlenecks.
• Interacting with architecture, software and hardware engineers to propose solutions.
• Developing studies to quantify performance benefits of features and algorithms.
• Developing tools to measure, communicate and improve GPU performance.

Skills

GPU architecture

Debug skills

Data analysis

Python

C++

Team collaboration

Communication

Hardware design

SystemVerilog

Simulation tools

Tools

SystemVerilog

Simulation/Emulation tools

Job description

Apple in the United Kingdom seeks a GPU Performance Analysis Engineer to help design and optimize next‑generation GPUs across Apple products. You will work with architecture, software, and hardware teams to diagnose bottlenecks and propose optimisations, developing study methods and measurement tools.

Responsibilities include validating GPU performance through tests, analysing real content, and quantifying benefits of hardware/software features.