Senior Software Engineer - GPU Kernel Authoring & Optimization

CoreWeave Senior Software Engineer - GPU Kernel Authoring & Optimization Reposted 23 Hours AgoSaved In-Office 2 Locations 182K-242K Annually Senior level 182K-242K Annually Senior levelCloud • Information Technology • Machine LearningAuthor, profile, and optimize CUDA GPU kernels for LLM inference to maximize throughput and minimize latency. Build reproducible microbenchmarks and MLPerf workflows, collaborate across product, hardware, and orchestration teams, lead designs and reviews, mentor engineers, and translate kernel wins into end-to-end model-serving performance improvements.Top Skills: C++CudaLlm-DMlperfNsight ComputeNsight SystemsNvlinkPciePythonSglangTensor CoresTensorrt-LlmVllm