Early Career - GPU Performance Architect

Summary

The hardest problems are the ones worth solving. We believe meaningful progress comes from pushing past where it’s easy to stop, and that a team with the right people and relentless drive can outperform even its own ambitions.

We’re looking for highly motivated, innovative engineers to join the GPU Performance Modelling Team. This group drives advanced exploration of Apple’s next-generation GPU architectures and micro-architectures. The role is highly visible and sits at the heart of the chip design effort, collaborating across every discipline involved. You’ll play a direct part in delivering high-performance products to millions of customers.

Description

Every generation of Apple silicon raises the bar for what a GPU and SoC can do: more performance, more efficiency, in tighter power envelopes than the industry thought possible. As a Performance Architect on this team, you’ll help decide where that bar goes next.

You’ll join a group shaping the architecture of Apple’s GPU and SoC from the ground up—researching ideas that become the next leap forward. The work is deeply collaborative: you’ll partner with software and hardware engineering teams to pressure-test ideas, refine promising directions, and seed new ones. You’ll investigate workloads that matter across today’s products and tomorrow’s categories to find the bottlenecks worth solving and the opportunities worth pursuing.

In this role, you will:

• Research and propose novel ideas that help shape Apple’s platform direction
• Prototype architectural and micro-architectural concepts in existing C++ performance models
• Gather, analyze, and validate measured and simulated results to identify bottlenecks and evaluate new ideas
• Present findings and recommendations to peers and leadership
• Collaborate with cross-functional teams to turn ideas into shipped architecture

If you’re energized by working at the edge of what silicon can do and want to see your ideas land in products used by hundreds of millions of people, we’d love to talk.

Minimum Qualifications

• Strong knowledge of computer architecture and micro-architecture (CPU, GPU, SoC, or adjacent domains)
• Hands-on experience with performance and/or power simulation environments
• Skilled at analyzing and characterizing complex workloads

Preferred Qualifications

• Familiarity with common data structures, algorithms, and design patterns
• Working knowledge of C++ and Python
• Comfortable working in uncertain environments and through ambiguity
• Strong communicator who can influence decisions and negotiate trade-offs