Senior Silicon Architect

Google Google · Big Tech · Mountain View, CA +1

This role is for a Senior Silicon Architect at Google DeepMind, focusing on defining chip IP architectures, memory hierarchy, and interconnect fabrics. While the role involves AI/ML accelerator architecture design as a qualification, the primary responsibilities are centered around silicon architecture and system-level optimization, not the direct development or deployment of AI models or agents.

What you'd actually do

  1. Lead the definition of chip IP architectures, memory hierarchy, interconnect fabrics, and IP integration to meet product requirements.
  2. Utilize high-level performance and power models to guide architectural decisions and conduct detailed trade-off analyses to optimize for system-level goals.
  3. Create and maintain detailed architectural specification documents that guide micro-architecture and design teams throughout the project lifecycle.
  4. Collaborate with software, product management, and physical design teams to ensure the hardware architecture is feasible, aligns with software needs, and meets all product requirements.
  5. Drive the technical evaluation, selection, and integration strategy for key third-party IP cores and technologies. Research emerging technologies and industry trends to influence and contribute to the long-term silicon and system architecture roadmap.

Skills

Required

  • Microarchitecture design (pipelines, caches, TLBs, execution systems, memory)
  • System-level optimization (performance, power, reliability, cost)
  • IP integration for System-on-Chip (SoC) designs
  • RTL code analysis (Verilog or VHDL)
  • Technical leadership (project planning, driving decisions)
  • Machine Learning (ML) accelerator architecture design

What the JD emphasized

  • 1 year of experience in Machine Learning (ML) accelerator architecture design