AI/ML Performance Engineer – Training on Trainium

Amazon Web Services (AWS); Cupertino (CA); On-site; USD 140,000 - 210,000; Full time

Benefits offered by this job

Health Insurance

Medical Insurance

Dental Insurance

Vision Insurance

Prescription Insurance

Basic Life & AD&D Insurance

Supplemental Life Plans

EAP

Mental Health Support

Medical Advice Line

Flexible Spending Accounts

Adoption and Surrogacy Reimbursement

401(k) Matching

Paid Time Off

Parental Leave

Sign-on Payments

Restricted Stock Units (RSUs)

Job summary

Amazon Web Services (AWS) is seeking a software professional to lead optimization of distributed training on AWS Trainium, aiming to maximize throughput and model FLOPs utilization. You will diagnose bottlenecks across the stack, from collective communications to compiler optimizations, and drive end-to-end improvements.

Requirements include 3+ years of software development experience and 2+ years in system design or architecture, with proficiency in at least one programming language.

Qualifications

• Requires 3+ years of professional software development experience and 2+ years of experience in system design or architecture.
• Proficiency in at least one programming language is required.
• Preference for candidates with a Bachelor's degree in Computer Science.

Responsibilities

• Lead efforts to optimize distributed training performance on AWS Trainium.
• Identify and resolve performance bottlenecks across the software stack, from collective communications to compiler optimizations.

Skills

Distributed Training

Performance Optimization

PyTorch

JAX

AWS Neuron

Machine Learning

Compiler Optimization

Kernel Performance

System Architecture

Software Development Life Cycle

Collective Communications

Memory Utilization

Education

Bachelor's degree in Computer Science

Job description

Amazon Web Services (AWS) is seeking a software professional to lead optimization of distributed training on AWS Trainium, aiming to maximize throughput and model FLOPs utilization. You will diagnose bottlenecks across the stack, from collective communications to compiler optimizations, and drive end-to-end improvements.

Requirements include 3+ years of software development experience and 2+ years in system design or architecture, with proficiency in at least one programming language.