Power Architect (AI Data Center Chiplets)
Мэтч & Сопровод
Для мэтча с этой вакансией нужен Plus
Описание вакансии
TL;DR
Power Architect (AI Data Center Chiplets): Designing and optimizing power management and SoC architecture for AI data center products with an accent on energy efficiency and RISC-V based CPUs. Focus on power delivery networks, thermal analysis, and maximizing performance trade-offs for cutting-edge AI hardware.
Location: Hybrid based out of Santa Clara, CA. Must be eligible to access U.S. export-controlled technology.
Salary: $100k - $500k
Company
is leading the industry in cutting-edge AI technology, specializing in high-performance RISC-V CPUs and comprehensive AI platforms.
What you will do
- Develop SoC power architecture and optimization strategies for AI data center products.
- Analyze AI and ML workloads to improve energy efficiency and performance.
- Design experiments and perform root cause analysis for power-related issues.
- Collaborate with multi-disciplinary teams across silicon, firmware, and systems.
- Contribute to Open Chiplet Architecture and industry standards like OCMI.
Requirements
- Deep expertise in Power Management, SoC Power Architecture, and Power Delivery Networks.
- Proficiency in Power Estimation Methodologies (e.g., PTPX) and trade-off modeling.
- Experience with silicon PVT corners, binning, and voltage droop mitigation techniques.
- Ability to perform thermal analysis and evaluate performance trade-offs.
- Eligibility to access U.S. export-controlled technology (EAR compliance).
Culture & Benefits
- Highly competitive compensation package and comprehensive benefits.
- Collaborative and curious environment focused on solving hard technical problems.
- Opportunity to work on industry-shaping AI hardware from chiplets to full systems.
- Fast-paced environment that values bold ideas and rapid execution.
Будьте осторожны: если работодатель просит войти в их систему, используя iCloud/Google, прислать код/пароль, запустить код/ПО, не делайте этого - это мошенники. Обязательно жмите "Пожаловаться" или пишите в поддержку. Подробнее в гайде →