Virtual Platform Engineering Sr. Manager, Annapurna Labs Machine Learning Accelerators, AWS

AWS's Trainium and Inferentia chips power the world's largest machine learning clusters. Our team builds virtual platforms — full-system C++ models of these custom SoCs — that let software teams start development months before silicon exists. For Trainium3, our virtual platform enabled a full training workload within 12 hours of first silicon, putting servers in customer hands within weeks!

We're hiring a hands-on engineering manager to lead and scale our virtual platform effort. You'll own the end-to-end delivery of virtual platforms that our software partners depend on — from model architecture and level of detail, through deployment and customer enablement. This is a builder-leader role: you'll set the technical direction while staying close to the code and your customers.

What you'll do:
- Lead the team delivering virtual platforms used by design verification as well as driver, runtime, collectives, and application software teams to develop and validate software pre-silicon
- Own the virtual platform roadmap — deciding what to model, at what fidelity, and when to deliver it, based on customer needs and chip schedules
- Drive platform usability, performance, and scalability so teams can run real workloads on your models efficiently
- Build and improve the tooling, CI, and release infrastructure around the virtual platform so customers get reliable, well-documented drops
- Partner closely with software teams and design verification to understand their workflows and shape the platform to maximize their productivity
- Hire and develop a team of strong modeling engineers, setting high standards for code quality, testing, and delivery
- Dive into technical problems when needed - debug model issues, review architecture decisions, and unblock the team

Why this role is interesting:
- Your virtual platform directly accelerates AWS's most strategic silicon programs — software teams literally can't start without you
- You'll own a product with real internal customers who give you direct feedback, not just a component buried in a larger system
- The problem space is rich: full-system simulation, multi-subsystem integration, QEMU development, performance at scale, machine learning at the bleeding edge
- Small team, big impact, startup pace inside AWS's custom silicon org

No ML background needed. You’ll learn the required ML domain knowledge on the job. What matters is deep virtual platform or system modeling experience and the ability to lead a technical team.

About the team
More details about Trainium3, our team's latest achievement, as well as some insights into our team culture:
- (https://www.aboutamazon.com/news/aws/trainium-3-ultraserver-faster-ai-training-lower-cost)