Senior Software Development Engineer, EC2 Nitro

EC2 Nitro drives the planet's largest, fastest growing and most feature-rich compute cloud. Nitro is AWS ground-up design for virtualization at global scale built on a fully custom stack of hardware, firmware and applications. Nitro has enabled EC2 to support Intel, AMD and Amazon's custom silicon - the Graviton processor family - while raising the industry bar for security and performance across our product line.

We integrate hardware, firmware, application software and services to deliver new virtualized and bare-metal compute platforms for companies from startups through the Fortune 500. We are looking for an experienced leader to drive software development and scaling for new EC2 compute platforms. In this role, you will work with a broad and deep group of technical teams that develop hardware, firmware, systems and application software.

The ideal candidate is expected to have solid understanding of computer science fundamentals, and expertise in C, C++ or Rust development in a Linux environment. Experience with Linux package management, version control systems, automated build processes, and software unit testing are required. In-depth knowledge of ML frameworks and cluster management is highly preferred.

Key job responsibilities

• Design and develop innovative technologies that power the infrastructure supporting machine learning workloads on Ultraservers
• Lead technical projects establishing EC2 as the pioneer in cloud computing for ML workloads across diverse applications including LLMs, multimodal systems, and emerging model architectures.
• Develop and maintain comprehensive regression testing systems that validate performance across major component releases including frameworks, firmware, drivers, and networking infrastructure.
• Collaborate with hardware engineering teams to influence future platform designs based on performance insights gathered from state-of-the-art research and customer workloads.
• Build customer relationships by investigating complex performance challenges, developing solutions, and publishing actionable best practices through multiple channels.

About the team

The EC2 Machine Learning Supercompute team develops next generation Ultraserver platforms to power high-performance training and inference workloads.