- Palo Alto, California, United States
- Full-time · Senior
- $220K–$357K per year
- 5+ years
Design, build, and optimize massive GPU clusters for extreme-scale training and inference workloads; develop and tune low-level CUDA kernels; profile, debug, and eliminate bottlenecks across GPU memory hierarchy, networking fabric, filesystems, and multi-GPU operation; collaborate closely with AI research teams to deliver production-grade performance and scalability.
Compensation
$220,000 - $357,000 USD
Responsibilities
- Design, build, and optimize massive GPU clusters for extreme-scale training and inference workloads
- Develop and tune low-level CUDA kernels (GeMM, Attention, etc.), using CUTLASS, Tensor Cores, and Nsight for maximum performance
- Profile, debug, and eliminate bottlenecks across GPU memory hierarchy, networking fabric, filesystems, and multi-GPU operation
- Collaborate closely with AI research teams to deliver production-grade performance and scalability
Requirements
- Nice to have: Deep low-level systems programming
- Nice to have: Strong experience with large-scale GPU clusters or distributed compute infrastructure at production scale
- Nice to have: Hands-on work with GPU kernel optimization
- Nice to have: Track record of building or running high-performance infrastructure for AI workloads
- Nice to have: Ability to reason from first principles and optimize for both memory-bound and compute-bound scenarios