Machine Learning - Compiler Engineer , AWS Neuron, Annapurna Labs

Annapurna Labs (U.S.) Inc.; Cupertino (CA); On-site; USD 180,000 - 240,000; Full time

Job summary

Annapurna Labs (U.S.) Inc. is seeking a software engineer for the AWS Neuron Compiler team. You will build and optimize a compiler that translates ML models from frameworks like PyTorch, TensorFlow, and JAX to run efficiently on AWS Inferentia and Trainium.

You will work with accelerators and contribute to pre-silicon design while collaborating with internal and external stakeholders. You should have strong experience in C++/Java, compiler development, and ML frameworks, with familiarity with

Qualifications

• Experience with object-oriented languages (C++/Java) is required.
• Experience with compilers or ML models on accelerators is preferred.
• Familiarity with ML frameworks (PyTorch, TensorFlow, JAX) is relevant.
• Knowledge of OpenXLA, StableHLO, MLIR is a plus.

Responsibilities

• Build next-generation Neuron compiler transforming ML models to AWS Inferentia/Trainium targets.
• Solve compiler optimization problems to maximize performance for large model families.
• Collaborate with customers/stakeholders to gather requirements and communicate tradeoffs.
• Participate in pre-silicon design and bring new features to market.
• Design and implement compiler passes and verification for accelerator tech.

Skills

C++

Java

Compiler design

ML frameworks

Tools

OpenXLA

StableHLO

MLIR

Job description

Do you want to be part of AI revolution? At Amazon our vision is to make deep learning pervasive for everyday developers and to democratize access to AI hardware and software infrastructure. In order to deliver on that vision, we've created innovative software and hardware solutions that make it possible. AWS Neuron is the SDK that optimizes the performance of complex ML models executed on AWS Inferentia and Trainium, our custom chips designed to accelerate deep-learning workloads.

Key job responsibilities

This role is for a software engineer in the Compiler team for AWS Neuron. As part of this role, you will be responsible for building next generation Neuron compiler which transforms ML models written in ML frameworks (e.g, PyTorch, TensorFlow, and JAX) to be deployed AWS Inferentia and Trainium based servers in the Amazon cloud. You will be responsible for solving hard compiler optimization problems to achieve optimum performance for variety of ML model families including massive scale large language models like Llama, Deepseek, and beyond as well as stable diffusion, vision transformers and multi-model models. You will be required to understand how these models work inside-out to make informed decisions on how to best coax the compiler to generate optimal implementation instruction. You will leverage your technical communications skill to partner with internal and external customers/stakeholders and will be involved in pre-silicon design, bringing new products/features to market, ultimately, making Neuron compiler highly performant and easy-to-use.

Experience in object-oriented languages like C++/Java is a must, experience with compilers or building ML models using ML frameworks on accelerators (e.g., GPUs) is preferred but not required. Experience with technologies like OpenXLA, StableHLO, MLIR will be added bonus!

Explore the product and our history! https://awsdocs-neuron.readthedocs-hosted.com/en/latest/neuron-guide/neuron-cc/indexhtmlhttps://aws.amazon.com/machine-learning/neuron/https://github.com/aws/aws-neuron-sdk https://www.amazon.science/how-silicon-innovation-became-the-secret-sauce-behind-awss-success

AWS Utility Computing (UC) provides product innovations - from foundational services such as Amazon's Simple Storage Service (S3) and Amazon Elastic Compute Cloud (EC2), to consistently released new product innovations that continue to set AWS's services and features apart in the industry. As a member of the UC organization, you'll support the development and management of Compute, Database, Storage, Internet of Things (Iot), Platform, and Productivity Apps services in AWS, including support for customers who require specialized security solutions for their cloud services.

A day in the life

As you design and code solutions to help our team drive efficiencies in compiler architecture, you'll create compiler optimization and verification passes, build features surface features and peculiarities of AWS accelerators to developers, implement tools to analyze numerical errors, and resolve the root cause of compiler defects. You'll also participate in design discussions, code review, and communicate with internal (other Neuron SDK and Amazon wide teams) and external stakeholders (open-source communities). Lastly, work in a startup-like development environment, where you're always working on the most important stuff.

About the team

Our team is dedicated to supporting new members. We have a broad mix of experience levels and tenures, and we're building an environment that celebrates knowledge-sharing and mentorship. Our senior members enjoy one-on-one mentoring and thorough, but kind, code reviews. We care about your career growth and strive to assign projects that help our team members develop your engineering expertise so you feel empowered to take on more complex tasks in the future.

Diverse Experiences

Amazon values diverse experiences. Even if you do not meet all of the qualifications and skills listed in the job description, we encourage candidates to apply. If your career is just starting, hasn't followed a traditional path, or includes alternative experiences, don't let it stop you from applying.