Staff AI Runtime Architect - Scale GPU Training

Databricks; Mountain View (CA); On-site; USD 190,000 - 265,000; Full time

Benefits offered by this job

Annual performance bonus

Equity options

Comprehensive benefits package

Job summary

Databricks is looking for a Staff Software Engineer for AI Runtime in Mountain View, California, to help build and scale the systems necessary for large-scale GPU training. You will drive the architecture of a managed GPU training platform ensuring high throughput and resilience across many accelerators.

The ideal candidate will have over 10 years of experience in distributed systems and machine learning, with a solid grasp of GPU performance. Join Databricks to make a significant impact in the AI training domain.

Qualifications

• 10+ years of experience in large-scale distributed systems.
• Hands-on experience with distributed training frameworks.
• Understanding of GPU performance fundamentals.

Responsibilities

• Drive the architecture of AIR’s managed GPU training platform.
• Solve large-scale training problems.
• Lead engineering efforts from design to production.

Skills

Large-scale distributed systems

GPU training infrastructure

Machine Learning systems

Distributed training frameworks

High-performance computing

Education

BS in Computer Science or related field

MS or PhD preferred

Tools

PyTorch

FSDP

DeepSpeed

Megatron

Job description

Databricks is looking for a Staff Software Engineer for AI Runtime in Mountain View, California, to help build and scale the systems necessary for large-scale GPU training. You will drive the architecture of a managed GPU training platform ensuring high throughput and resilience across many accelerators.

The ideal candidate will have over 10 years of experience in distributed systems and machine learning, with a solid grasp of GPU performance. Join Databricks to make a significant impact in the AI training domain.