AI Systems Engineer, Hardware Architecture

Meta Reality Labs is seeking a principal-level AI Systems Engineer to define the hardware architecture strategy for next-generation AI-accelerated computing systems powering virtual and augmented reality products. In this role, you will shape the long-term silicon and systems roadmap for on-device AI inference and training workloads across wearables, headsets, and spatial computing platforms. You will drive architectural decisions that span custom silicon, memory subsystems, interconnects, and software-hardware co-design, ensuring Meta's AI hardware remains at the forefront of performance, efficiency, and capability for immersive computing experiences.

Responsibilities

Define and own the multi-year architectural roadmap for AI compute subsystems across Meta's hardware product lines, including wearables and spatial computing devices Lead system-level architecture exploration for AI inference and training accelerators, including memory hierarchy design, interconnect topology, and power-performance-area tradeoffs Drive cross-functional alignment across silicon engineering, firmware, software, and product teams to translate AI workload requirements into hardware specifications Develop and maintain architectural models, performance simulators, and analytical frameworks to evaluate design tradeoffs at the system level Identify and resolve architectural bottlenecks across the AI compute stack, from neural network operators to silicon microarchitecture Establish technical direction for AI hardware platform decisions, including custom silicon versus third-party IP evaluation and integration strategies Partner with machine learning researchers and compiler teams to co-design hardware-software interfaces that maximize AI model efficiency on constrained wearable platforms Represent hardware architecture in executive-level technical reviews, authoring detailed architecture decision records and system specifications Mentor and provide technical guidance to other engineers across hardware architecture and systems engineering disciplines Evaluate emerging AI workloads, model architectures, and compute paradigms to proactively inform future hardware platform investments

Qualifications

15+ years of experience in hardware systems architecture, with a focus on AI, ML, or high-performance compute systems Experience defining SoC or system-level architecture for AI inference or training workloads, including memory subsystem design, compute hierarchy, and interconnect topology Experience with hardware-software co-design methodologies for on-device AI workloads, including familiarity with ML compiler stacks, operator fusion, and quantization impacts on hardware design Experience developing system performance models and using simulation or analytical frameworks to evaluate architectural trade-offs at scale Track record of driving multi-year hardware architecture roadmaps and influencing silicon strategy across large engineering organizations Familiarity with custom silicon development flows, including architecture-to-RTL handoff, physical design constraints, and post-silicon validation feedback loops Experience architecting AI systems for power- and area-constrained wearable or mobile devices, including VR headsets, AR glasses, or similar spatial computing platforms Experience in evaluating and integrating emerging memory technologies (e.g., HBM, LPDDR5X, in-memory compute) into AI system architectures Background in collaborating with ML research teams to translate novel model architectures into hardware-efficient deployment targets