Overview
Join AWS Neuron as a Software Engineer on the Compiler team. You will help build the next generation Neuron compiler to transform ML models from frameworks (e.g., PyTorch, TensorFlow, JAX) for deployment on AWS Inferentia and Trainium. You will work to optimize performance across model families including large language models, Stable Diffusion, vision transformers, and multi-model workloads. You will collaborate with customers and internal teams, contribute to pre-silicon design, and help bring new features to market to make the Neuron compiler highly performant and user-friendly.
Key responsibilities
* Design, implement, test, deploy, and maintain software to improve Neuron compiler performance, stability, and user experience.
* Collaborate with chip architects, runtime/OS engineers, scientists, and ML Apps teams to deploy models on AWS accelerators with optimal cost/performance.
* Work with open-source components (e.g., StableHLO, OpenXLA, MLIR) to optimize ML workloads on AWS software and hardware.
* Develop features that deliver excellent experiences for developers worldwide.
A day in the life
Design and code compiler optimizations and verification passes; surface accelerators’ peculiarities to developers; build tools to analyze numerical errors; diagnose and fix compiler defects; participate in design discussions and code reviews; engage with internal and external stakeholders; operate in a startup-like development pace.
Basic Qualifications
* 3+ years of non-internship professional software development experience
* 2+ years of non-internship design or architecture experience (design patterns, reliability, scaling)
* Experience programming with at least one software programming language
Preferred Qualifications
* Master’s degree or PhD in Computer Science or related field
* 3+ years of production-grade code experience in object-oriented languages (C++, Java)
* Experience in compiler design for CPU/GPU/ML accelerators
* Experience with OpenSource compiler toolsets like LLVM/MLIR
* Experience with PyTorch, OpenXLA, StableHLO, JAX, TVM, DL models and related algorithms
* Experience with modern build systems (Bazel, CMake)
Amazon is an equal opportunity employer and does not discriminate on the basis of protected veteran status, disability, or other legally protected status.
Our inclusive culture empowers Amazonians to deliver the best results for our customers. If you have a disability and need a workplace accommodation or adjustment during the application and hiring process, please visit Amazon’s accommodations information for more details.
Our compensation reflects the cost of labor across US geographic markets. The base pay range for this position is provided to reflect typical market variations and may differ based on location, knowledge, skills, and experience. This position may include additional compensation elements as part of a total package.
This position will remain posted until filled. Applicants should apply via our internal or external career site.
#J-18808-Ljbffr