Senior Software Engineer - Hardware Abstraction Layer, Team Lead, AWS Machine Learning Accelerators

AWS's Annapurna Labs is seeking a hands-on technical leader for their system-on-chip hardware abstraction layer (SoC HAL) software team. This role is critical in managing AWS machine learning servers (Trainium and Inferentia) through backend software development. The position involves leading a small team of 3-5 developers while working closely with hardware designers and system software teams to build and maintain HALs for newly developed SoC IPs.

The ideal candidate will have strong C++ programming skills, experience with hardware-based systems, and the ability to solve complex system-level issues. While the team works on ML chips, no machine learning background is required as the focus is on low-level infrastructure management.

This is an opportunity to work with AWS's custom silicon division, Annapurna Labs, which designs and deploys some of the largest custom silicon in the world. The role offers the chance to work in a startup-like environment within AWS, focusing on innovation and technical excellence.

The position can be based in either Cupertino, CA or Austin, TX, with competitive compensation ranging from $151,300 to $261,500 based on location and experience. AWS offers a comprehensive benefits package including equity, sign-on payments, and various medical and financial benefits.

The role requires 6+ years of programming experience, strong system architecture skills, and experience with low-level software development. You'll be part of AWS Utility Computing (UC), supporting the development of various AWS services while working on cutting-edge cloud computing solutions.

AWS values diversity and inclusion, offering various employee-led affinity groups and ongoing learning experiences. The company emphasizes work-life harmony and provides extensive opportunities for mentorship and career growth. Join AWS to be part of a team that's changing the industry and building the future of machine learning infrastructure.