Machine Learning Compiler Architect

The Product: AWS Machine Learning accelerators are leading the charge in innovation on AWS, featuring tools designed for building cutting-edge Generative AI. Our Inferentia chip is engineered for unparalleled ML inference performance at the lowest cloud cost, while Trainium aims to deliver exceptional ML training capabilities with an industry-leading teraflops (TFLOPS) of compute power. This is all supported by our advanced AWS Neuron Software Development Kit (SDK), which consists of an ML compiler and runtime that seamlessly integrates with popular frameworks like PyTorch, TensorFlow, and MxNet. Organizations such as Snap, Autodesk, Amazon Alexa, and Amazon Rekognition rely on AWS Neuron and Inferentia to optimize performance. The Team: The Annapurna Labs team at Amazon is pivotal in silicon development for AWS, incorporating various disciplines such as silicon engineering, hardware design, software, and operations. The AWS Neuron team specializes in optimizing complex neural net models on custom-built AWS hardware. Our focus is on developing an advanced deep learning compiler stack that converts neural network descriptions from frameworks like TensorFlow, PyTorch, and MXNET into executable code. This ambitious project demands innovative solutions from some of the brightest engineering and research minds in the industry. You: We are searching for a skilled Machine Learning Compiler Engineer II to join our AWS Neuron team. In this role, you will contribute to the development and scaling of a compiler designed to manage significant ML workloads. Your responsibilities will include architecting and implementing essential features, conducting cutting-edge research, and collaborating closely with AWS ML services teams. Participation in pre-silicon designs and the rollout of exciting new products/features will also be part of your journey. While a background in Machine Learning and AI accelerators is preferred, it is not a strict requirement. About Us: At AWS, we celebrate diversity and are committed to cultivating an inclusive work environment. Our team consists of various employee-led affinity groups, fostering a culture where every voice is heard. We prioritize work-life balance by offering flexible working hours and encouraging a healthy blend of personal and professional life. We value mentorship and career growth, ensuring our team members have opportunities for development and complex project involvement. BASIC QUALIFICATIONS 3+ years of professional software development experience (non-internship) 2+ years of experience in architecting and optimizing compilers Proficiency in at least one programming language: C++ (preferred), C, or Python PREFERRED QUALIFICATIONS M.S. or Ph.D. in Computer Science or a related field Experience with multiple toolchains and Instruction Set Architectures Proficiency in resource management, scheduling, code generation, and compute graph optimization Experience in optimizing deep learning models in TensorFlow, PyTorch, or MxNET Amazon is dedicated to equal opportunity employment and prohibits discrimination based on protected veteran status, disability, or any other legally protected status. Our inclusive culture empowers Amazonians to deliver the best results for our customers. Location: This position will remain posted until filled. Applicants should apply via our internal or external career site.