GenAI Accelerator Architect - High-Performance Silicon

MatX; Mountain View (CA); Hybrid; USD 120,000 - 600,000; Full time

Benefits offered by this job

4 weeks PTO

Remote work up to 3 weeks

Medical insurance

401K

Job summary

MatX, based in Mountain View, CA, designs AI silicon and GenAI accelerators, delivering high-performance compute for the world’s leading models. We are seeking an AI Accelerator Compute Architect to define the compute architecture for our next‑generation accelerators, from ISA down to interconnect.

You will collaborate with Research, SW, and HW teams to optimize latency, throughput, and power, evaluate tradeoffs, and support silicon bring‑up.

Qualifications

• Background in computer architecture and AI accelerators.
• Experience translating workloads into hardware designs.
• Strong microarchitecture fundamentals and latency/throughput knowledge.
• Experience with performance analysis and optimization.

Responsibilities

• Define AI compute subsystem architecture (ISA, memory, interconnect).
• Derive requirements from AI use cases and workloads.
• Collaborate with Research, SW, HW to balance performance and programmability.
• Lead power/area/performance tradeoffs with stakeholders.
• Support first-silicon bring-up and post-silicon debugging.

Job description

MatX, based in Mountain View, CA, designs AI silicon and GenAI accelerators, delivering high-performance compute for the world’s leading models. We are seeking an AI Accelerator Compute Architect to define the compute architecture for our next‑generation accelerators, from ISA down to interconnect.

You will collaborate with Research, SW, and HW teams to optimize latency, throughput, and power, evaluate tradeoffs, and support silicon bring‑up.