Amazon
Senior Power and Performance Architect, Annapurna Labs, Machine Learning Hardware
$183K–$248K
Cupertino, CA · Senior · Full-time · Posted Jan 28, 2026
- Experience
- 3+ years required; 10+ years preferred for power architecture in high-performance computing systems.
- Education
- Bachelor's degree in Electrical Engineering, Computer Engineering, or a related technical field; master's degree or equivalent preferred.
About the job
Develop power architecture and management solutions for machine-learning accelerator platforms, optimizing power, performance, and thermal constraints in Amazon data centers.
What you'll do
- Design power architecture, DVFS flows, control schemes, and power-management algorithms for ML accelerators.
- Lead peak-power and di/dt management; analyze workloads and power-performance trade-offs.
- Collaborate across hardware and software teams, build power models and analysis scripts, and provide technical leadership and mentoring.
What they ask for
- Bachelor's degree in Electrical Engineering, Computer Engineering, or a related technical field.
- 3+ years of experience in power architecture and management.
- Strong background in DVFS and power optimization; experience designing power control schemes and algorithms, including peak-power management and di/dt handling.
- Python proficiency; workload optimization experience; working knowledge of C/C++ and software development lifecycle; understanding of RTL design and pre-silicon validation.
Nice to have
- Master's degree or equivalent; 10+ years of experience in power architecture for high-performance computing systems.
- Experience with machine learning accelerator architectures, power modeling and analysis tools, or data center power management solutions.
Pay
183,000.00 - 247,600.00 USD annually (base salary)
Tools
DVFS, Python, C/C++, RTL
AI focus
Power architecture and optimization for machine-learning accelerator platforms.
Model-written summary · the original posting is authoritative