Description
The company is hiring GPU Performance / Kernel Engineers at Bellevue, Washington, in a hybrid role requiring approximately three days per week in the office. The role focuses on optimizing GPU utilization, latency, and throughput for large-scale AI training and inference workloads by tuning kernels, profiling and debugging performance bottlenecks, developing benchmarking practices, and evaluating GPU technologies. Candidates should have strong GPU kernel development and performance optimization experience, preferably with CUDA or ROCm, and the ability to work across distributed AI and cloud GPU infrastructure.
