ML Systems Engineer - Distributed Training on Trainium

Socket.dev

Seattle (WA)

On-site

USD 144,000 - 194,000

Full time

8 days ago

Get more replies from employers

Send a job-specific resume in minutes.

Job summary

Amazon Web Services (AWS) is hiring for a role within the Annapurna Labs team that builds AWS Neuron to accelerate deep learning and GenAI workloads on AWS Trainium. You will optimize distributed training across the Neuron software stack, collaborating with PyTorch and JAX teams to enable large-scale training on the latest Trainium instances.

The role involves leading parallelism strategies, profiling workloads, and driving fixes across compiler, runtime, and collectives teams to improve

Qualifications

3+ years of non-internship professional software development experience.
2+ years of non-internship design or architecture (design patterns, reliability and scaling) of new and existing systems experience
Experience programming with at least one software programming language.
3+ years of full software development life cycle, including coding standards, code reviews, source control management, build processes, testing, and operations experience
Bachelor's degree in computer science or equivalent

Responsibilities

Lead efforts to optimize distributed training performance on Trainium across the Neuron software stack.
Own parallelism strategies (data, tensor, pipeline, expert, context) and apply reduced-precision formats when beneficial.
Profile workloads end-to-end to identify bottlenecks and drive fixes with compiler, runtime, and collectives teams.
Translate performance gaps into framework requirements and contribute upstream to open source frameworks.