Description
AWS Infrastructure Services owns the design, planning, delivery, and operation of all AWS global infrastructure. We're seeking a customer-obsessed, team-driven technology leader to take our engineering of server hardware and software to the next level.
As a server performance leader, you'll work on Amazon's hardest problems, building high-quality, architecturally sound systems that are aligned with AWS business needs. You'll interact with internal and external interdisciplinary teams of engineers to design, develop, validate, launch, monitor, and troubleshoot at large scale, ensuring servers are operating at maximum performance while staying within thermal budgets.
You'll work closely with CPU/GPU suppliers to align and influence roadmaps, driving future architectures. This is a fast-paced, intellectually challenging position.
Responsibilities
- Design, develop, validate, launch, monitor, and troubleshoot large-scale server systems
- Collaborate with internal and external teams to ensure maximum performance and thermal efficiency
- Work with CPU/GPU suppliers to influence roadmaps and drive future architectures
Requirements
- Bachelor's degree in Computer Engineering, Electrical Engineering, Electronics Engineering, or related discipline
- Experience in developing functional specifications, design verification plans, and functional test procedures
- 5+ years of experience identifying and root-causing HW/SW issues in large-scale systems
- 5+ years of experience with customer support for successive debugging/root-causing CPU performance issues
- 5+ years of experience in CPU/memory architecture and subsystem design
Preferred Qualifications
- Master's degree in electrical engineering, computer engineering, or equivalent
- 5+ years of experience with new CPU technology bring-up
- 5+ years of experience with statistical analysis using large-scale data
- 5+ years of experience with leading memory standardization through JEDEC, CPU architecture, RAS, DDR Memory technology, system performance, security, and troubleshooting