Principal NPU Microarchitect
Core
Lead the definition of Rivian's next-generation Neural Network Accelerator Engine (NPU) microarchitecture for production deep learning workloads in electric vehicles.
Role type
Principal NPU Microarchitect (Silicon/Hardware IC)
Builds
Custom ASIC/SoC compute engines (NPUs) for AI inference in automotive platforms
Domain
Automotive semiconductor / AI hardware acceleration
Deliverable
production ML models
Required skills
Computer architecture, logic design, silicon microarchitecture, NPU/TPU/AI accelerator design, systolic arrays, hardware scheduling, quantization-aware execution units, on-chip memory hierarchy, high-throughput DMA, hardware-software co-design, performance modeling, PPA optimization
Preferred skills
Automotive SoC/safety-critical silicon (ISO 26262), mapping Transformers/CNNs/LLMs to custom hardware, instruction-driven accelerator architectures, hardware modeling frameworks (SystemC, C++)
Technologies
SystemC, C++, Python, RTL, SoC
Responsibilities
Define and evolve NPU core microarchitecture including compute datapaths and hardware scheduling; Architect solutions mapping to custom hardware models including processing arrays and silicon partitioning; Drive architectural tradeoffs across PPA, utilization, and scalability; Lead definition of mechanisms for efficient tensor movement via on-chip/off-chip memory and DMA; Partner with compiler, firmware, and verification teams on hardware-software co-design; Build cycle-accurate performance models to guide design decisions; Mentor engineers across architecture and implementation disciplines
Seniority
Principal, hands-on IC with mentorship