Define and drive the end-to-end architecture of next-generation XPUs for AI infrastructure, focusing on scalability and performance. This includes designing compute, memory, and IO subsystems while collaborating with design and DV teams for implementation.
Role Overview
We are seeking aSoC Architectto define and drive the architecture of next-generation XPUs forAI Infrastructure. This role involves end-to-end ownership of system and SoC architecture, fromconcept to silicon, with a strong emphasis onscalability, performance, interconnects, and system-level optimization.
Key Responsibilities
Definesystem architecturefor complex SoCs, including compute, interconnect, memory, and IO subsystems
Drive architecture forscalable and modular designs, including chiplet-based systems (UCIe)
Architect high-performance data movement across:
Compute engines (ARM/RISC-V/ML/MAC)
Memory subsystems (DDR/LPDDR/HBM)
IO interfaces (PCIe/UCIe/CXL/UALink/ESUN)
Collaborate with design and DV teams to ensurearchitectural feasibility and efficient implementation
Developperformance models and simulationstovalidatearchitecture choices
Drive trade-offs across: Performance, Power, Area and Cost
Definesystem-level verification and validation strategies
Work closely with customers and software teams to align architecture withreal-world workloads
Influence long-term roadmap and technology direction
Required Qualifications
Bachelor’s/Master’s/PhD in Electrical Engineering, Computer Engineering, or related field
18+ years of experience inSoC or system architecture
Proventrack recordof architectingcomplex SoCs for AI, HPC, or Networking
Deep understanding of:
System-level design and trade-offs
High-speed IO protocols (UCIe, PCIe, Ethernet)
Data movement and interconnect design
Experience withperformance modeling and architectural simulation
“I was the first applicant for a remote marketing position that got listed on the company website the same day I applied. Had an interview within 48 hours!”