AI Systems Engineer, Hardware Architecture
Meta PlatformsJob Title
AI Systems Engineer, Hardware Architecture
Role Summary
Principal-level hardware architect responsible for defining long-term silicon and system architecture for on-device AI inference and training across AR/VR and wearable platforms. Primary focus areas include imaging systems, camera pipelines, custom AI accelerators, memory and interconnect subsystems, and hardware-software co-design to meet performance, latency, and efficiency targets for spatial computing products.
Experience Level
Senior - principal-level role. Requires substantial experience (see Requirements); the posting specifies 12+ years in hardware systems architecture.
Responsibilities
Lead system- and product-level architecture for AI and imaging in spatial computing devices. Key responsibilities include:
- Define multi-generation hardware architecture strategy for on-device AI inference and training across VR, AR, and wearables.
- Lead architectural exploration and trade-off analysis across compute, memory hierarchy, interconnects, and power delivery.
- Architect end-to-end camera and imaging pipelines, including sensor interfaces, ISP integration, and real-time image processing for perception and vision tasks.
- Drive hardware-software co-design with silicon, firmware, camera systems, and ML platform teams to optimize end-to-end performance.
- Specify camera subsystem requirements: multi-camera synchronization, depth sensing, low-latency visual processing, and related interfaces.
- Establish requirements and performance targets for custom AI accelerators, ISPs, SoCs, and supporting subsystems.
- Develop and maintain system performance models and simulation frameworks to evaluate architectural decisions against real-world workloads.
- Provide architectural guidance to hardware teams and communicate technical rationale to executive and cross-functional stakeholders.
- Engage external silicon partners, camera module vendors, and research groups to evaluate and incorporate emerging imaging technologies.
Requirements
Primary qualifications and technical skills required; preferred items listed separately.
Must-have:
- 12+ years of experience in hardware systems architecture, focused on AI, ML, imaging, or high-performance compute systems.
- Deep expertise in camera pipeline architecture, including image signal processing (ISP) and sensor integration.
- Experience defining SoC or system-level architecture for AI inference or training workloads, including memory subsystem design, compute hierarchy, and interconnect topology.
- Experience architecting imaging subsystems for real-time computer vision: multi-camera systems, depth sensing, and visual-inertial odometry.
- Experience with hardware-software co-design for on-device AI and imaging, and familiarity with ML compiler stacks and ISP tuning workflows.
- Experience developing system performance models and using simulation/analytical frameworks to evaluate architectural trade-offs.
- Proven track record driving multi-year hardware architecture roadmaps and influencing silicon strategy across large organizations.
Nice-to-have:
- Familiarity with custom silicon development flows: architecture-to-RTL handoff, physical design constraints, post-silicon validation.
- Experience targeting power- and area-constrained wearable or mobile devices (VR headsets, AR glasses, spatial computing platforms).
- Expertise in computational photography, HDR processing, or neural ISP architectures.
- Experience evaluating/integrating emerging memory technologies (e.g., HBM, LPDDR5X, in-memory compute).
- History collaborating with ML research and camera teams to translate algorithms into hardware-efficient implementations.
Education Requirements
Not specified.
About the Company
Company: Meta Platforms
Headquarters: Menlo Park, California, United States
American technology company that develops social networking products (Facebook, Instagram, WhatsApp) and invests in virtual/augmented reality hardware and software through Reality Labs, focusing on connectivity, advertising, and immersive computing experiences.
