We are forming a small, agile engineering team to accelerate a new initiative focused on artificial intelligence (AI) performance - from micro-architecture exploration through end-to-end workload validation. You’ll work at the intersection of silicon, systems, and software, partnering cross-functionally with chip and system architects, as well as inference software engineers to drive data-backed design decisions and deliver step-function improvements in throughput, latency, and efficiency. As a Principal Silicon Performance Architect on this AI acceleration effort, you will own performance modeling and analysis for current and future AI workloads. You will translate hardware and software architectural ideas into simulator implementations, run rigorous experiments across design variants, and turn results into clear guidance for architecture and product decisions.

• Extend and adapt simulation infrastructure to model new micro-architecture innovations for AI inference. • Analyze performance for current and forward-looking AI inference workloads across latency, throughput, and efficiency dimensions. • Drive design-space exploration using AI-assisted workflows, automation, and large-scale experiment generation. • Communicate performance insights clearly and influence architecture decisions through data-driven recommendations. • Collaborate closely with chip, system, and software architects to propose, evaluate, and iterate on architectural variations.