Software Engineer - Kernels/CUDA (C++)

Location: Palo Alto, California, or Seattle, Washington, United States

Salary: $180,000–$440,000 USD per year

Company

Builds AI systems and a large-scale AI supercomputer to support advanced training and inference workloads.

What you will do

• Design, build, and optimize massive GPU clusters for extreme-scale training and inference.
• Develop and tune low-level CUDA kernels, including GeMM and attention kernels, using CUTLASS, Tensor Cores, and Nsight.
• Profile and debug performance across GPU memory hierarchies, networking fabrics, filesystems, and multi-GPU systems.
• Build and improve the platform layer for compute infrastructure, including large-scale orchestration and virtualization.
• Collaborate with AI research teams to deliver production-grade performance and scalability.

Requirements

• Deep low-level systems programming experience with C, C++, PTX, or SASS.
• Experience with large-scale GPU clusters or distributed compute infrastructure in production.
• Hands-on experience optimizing GPU kernels with CUTLASS, custom kernels, and Nsight profiling.
• Experience building or operating high-performance infrastructure for AI training or inference.
• Ability to optimize memory-bound and compute-bound workloads from first principles.
• Strong written and verbal communication skills, with a hands-on engineering approach.

Culture & Benefits

• Flat organizational structure with direct contribution expected from all employees.
• Small, highly motivated team focused on engineering excellence and curiosity.
• Medical, vision, and dental coverage.
• 401(k) retirement plan, short- and long-term disability insurance, and life insurance.
• Equity, discounts, and additional benefits.