Software Engineer - GPU Kernels

About the role

GPU kernel engineering for mission-critical AI inference. Requires strong GPU architecture understanding, C++ proficiency, CUDA C++ API knowledge, and familiarity with quantization and modern GPU features such as tensor cores.

Requirements

• Strong understanding of GPU architecture: memory hierarchy, thread/block/grid organization, synchronization
• Proficient in C++ and GPU performance profiling tools
• Knowledge of CUDA C++ API, memory access patterns, bandwidth optimization, numerical precision and quantization
• Nice to have: Transformer/attention optimization, Cutlass/Triton/Thrust/CUB, GEMM tuning, multi-GPU compute

Where this was posted

• LinkedIn

  last seen 2mo ago
  View on LinkedIn