Lead performance optimization for large language model training and inference on NVIDIA GPU clusters, improving GPU utilization, memory footprint, and distributed training efficiency.
Locations (2)
Cambridge, Massachusetts, United States
San Francisco, California, United States
Compensation
Expected Base Salary Range
$148,000—$240,000 USD
Responsibilities
Profile and optimize GPU compute kernels for large language models
Optimize GPU cluster performance for collective/MoE frameworks and RDMA/IB/EFA usage
Manage big data pipelines for mid-training scale data
Optimize reinforcement learning pipelines and async cycles
Ensure optimized configurations are tested, accessible, and adaptable for research tasks
Requirements
Bachelor's or Master's in Computer Science or related quantitative field
PhD in a relevant research area
Nice to have: Kernel DSLs like CUDA, CuTe, CuTile, Triton
Nice to have: GPU networking frameworks like NCCL, NIXL, NVLink, IB, EFA
Nice to have: Distributed frameworks such as Ray, Flyte, Airflow
Design, implement, and maintain low-level driver abstractions and communication primitives for multi-accelerator systems.
Locations (3)
Los Altos, California, United States
United States
Canada
Compensation
The estimated base salary range for this role to be performed in the US, regardless of the state, is $148,000.00–$270,000.00 USD. The estimated base salary range for this role to be performed in Canada, regardless of the province, is $172,000.00–$258,600.00 CAD.
Responsibilities
Implement and extend core driver abstractions across diverse hardware backends
Contribute to multi-accelerator and multi-node communication and collectives primitives
Improve diagnostics and error reporting across asynchronous execution stack
Partner with peer teams to build and refine surfaces they consume
