Neuron Collectives Software Engineer, Trainium Collectives
Employment type
Full-time
Work setting
On-site
Location
Cupertino, CA
Schedule
Day shift
Posted
You'll be redirected to the employer's application page.
Job overview
The Neuron Collectives Software Engineer role is an onsite position based in Cupertino, CA. Compensation is not specified in the posting, and this role supports the Annapurna Labs team in building networking solutions for ML and HPC workloads. This role exists to enhance collective algorithms and topologies for optimal training performance. The engineer will optimize collective operations to scale AI compute across the data center.
What you'll do
- Enhance collective algorithms and topologies for optimal training performance.
- Use tools to identify bottlenecks in compute and bus bandwidth.
- Monitor and analyze processor, DMA, and firmware metrics.
- Optimize collective operations to scale AI compute.
- Develop and optimize C/C++ implementations of collective communication patterns.
What we're looking for
- Skills & competencies
- full-timecupertino casoftware engineeringmachine learningnetworkinghpccollectivesc++optimizationonsiteno on-callno travel
- Work arrangement
- Weekend coverage required
Benefits & perks
- Health insurance, 401(k) matching, paid time off, parental leave, sign-on payments, and RSUs.
Why this role
Focus on collective communication primitives for large-scale AI training.
About the employer
Annapurna Labs (U.S.) Inc. is hiring for this role. Industry: Computing Infrastructure Providers, Data Processing, Web Hosting, and Related Services. Sector: 51.
Additional details
- Industry sector
- 51
- Industry
- Computing Infrastructure Providers, Data Processing, Web Hosting, and Related Services
- Occupation code
- 15-1252.00
You'll be redirected to the employer's application page.
Listing ID: 9c48505d-08e1-4abc-8277-2a0cd620e37b