Senior CPU Performance Architect - Silicon Power

Google; Mountain View (CA); On-site; USD 163,000 - 237,000; Full time

Job summary

Google in Mountain View, CA is seeking a CPU Performance Architect to push boundaries in silicon design. You will lead deep-dive analyses, develop models, and collaborate with CPU architects to optimize performance and power across diverse workloads.

Qualifications include a Bachelor's in EE/CE/CS, 8 years in performance analysis, and proficiency with C/C++ and Python. This role offers impact on next-generation hardware and equity benefits.

Qualifications

• Bachelor's degree in EE/CE/CS or equivalent practical experience.
• 8 years of experience with performance analysis, microarchitectural bottleneck isolation, and workload characterization.
• Experience with C/C++ for architectural modeling and Python for data analysis and automation frameworks.

Responsibilities

• Lead deep-dive microarchitectural analysis on silicon to isolate and quantify pipeline stalls, memory hierarchy bottlenecks, and instruction-level inefficiencies to drive root-cause resolution.
• Design, build, and execute advanced power and performance experiments on physical silicon to rigorously correlate pre-silicon projections with post-silicon reality across general-purpose and AI/ML compute workloads.
• Develop and maintain early-stage power and performance models to evaluate complex architectural 'what-if' scenarios, driving hardware-software co-optimization.
• Partner closely with lead CPU architects and cross-functional teams to define dynamic, realistic power and performance goals for future Google silicon.

Skills

Performance analysis

Microarchitectural bottleneck

Workload characterization

C/C++

Python

Architectural modeling

Education

Bachelor's degree or equivalent

Job description

Google in Mountain View, CA is seeking a CPU Performance Architect to push boundaries in silicon design. You will lead deep-dive analyses, develop models, and collaborate with CPU architects to optimize performance and power across diverse workloads.

Qualifications include a Bachelor's in EE/CE/CS, 8 years in performance analysis, and proficiency with C/C++ and Python. This role offers impact on next-generation hardware and equity benefits.