Principal SoC Architect (AI)

Location: Santa Clara County, California, United States

Salary: $300,000–$350,000 per year

Company

Early-stage AI infrastructure startup backed by Tier-1 venture capital firms, developing chips and systems for highly efficient inference workloads.

What you will do

• Lead end-to-end SoC architecture for next-generation AI inference chips.
• Define key SoC blocks, interfaces, and architectural trade-offs across performance, power, area, schedule, and cost.
• Architect PCIe Gen5-or-higher connectivity, scale-up and scale-out interconnects, and UCIe integration.
• Design HBM and LPDDR memory subsystems, including bandwidth modeling and memory controller integration.
• Define CPU complex, boot flows, security, power, reset, and clocking control paths.
• Coordinate with RTL, firmware, validation, packaging, and bring-up teams through silicon debug and performance tuning.

Requirements

• Experience architecting real, shipping SoCs across the full development lifecycle.
• Strong knowledge of SoC blocks and their interfaces, including architecture, RTL, firmware, and bring-up workflows.
• Experience with PCIe Gen5 or higher, scale-up or scale-out interconnects, and UCIe.
• Experience with HBM or LPDDR memory architecture, bandwidth modeling, and memory controller integration.
• Experience designing CPU control paths, boot flows, security, power, reset, and clocking systems.
• Experience with silicon bring-up, debug strategy, and performance optimization.

Nice to have

• Early-stage SoC development experience in a startup environment.
• Chiplet or multi-die SoC experience.
• Experience with AI accelerators, GPUs, or high-bandwidth networking silicon.

Culture & Benefits

• Opportunity to make a substantial impact on a fundamentally new chip and system architecture.
• Work on AI infrastructure innovation at both the chip and system levels.
• Build products from the ground up in an early-stage startup environment.