SoC Performance Architect for AI Wearables

Meta; Austin (TX); On-site; USD 178,000 - 250,000; Full time

Benefits offered by this job

Bonus

Equity

Benefits

Job summary

Meta Reality Labs in Austin, TX seeks a Silicon Performance Architect to lead performance modeling, workload profiling, and HW-SW co-design efforts for advanced AI-enabled wearables. You will model IP and DRAM, analyze workloads, and guide architecture decisions with cross-functional teams.

The role requires 8+ years in performance modeling using C/C++ or SystemC, strong SoC/NoC knowledge, Python scripting, and a background in power trade-offs.

Qualifications

• Bachelor's degree in CS/CE or equivalent practical experience.
• 8+ years creating performance models with C/C++ or SystemC.
• Experience with SoC architecture, NoCs, QoS, memory sub-system.
• Python scripting, automation, and visualization.
• Understanding power concepts and trade-offs.
• Ability to document analyses for cross-functional teams.

Responsibilities

• Conduct bottleneck analysis of existing and future SoCs to identify improvements.
• Develop accurate models of IP components, SoC fabric, and DRAM to predict performance.
• Analyze workloads and create micro-benchmarks to evaluate performance.
• Identify optimization opportunities and collaborate on HW-SW co-design.
• Collaborate across disciplines, vendors, brainstorm ideas and coordinate multiple initiatives.

Skills

Performance modelling

C/C++

SystemC

SoC architecture

Python scripting

Power optimization

Data analysis

HW-SW co-design

Bottleneck analysis

Education

Bachelor's degree in CS/CE

MS in EE/CS

Tools

RTL verification

Job description

Meta Reality Labs in Austin, TX seeks a Silicon Performance Architect to lead performance modeling, workload profiling, and HW-SW co-design efforts for advanced AI-enabled wearables. You will model IP and DRAM, analyze workloads, and guide architecture decisions with cross-functional teams.

The role requires 8+ years in performance modeling using C/C++ or SystemC, strong SoC/NoC knowledge, Python scripting, and a background in power trade-offs.