Description
NVIDIA is hiring a Software Engineer to design, implement, and optimize high-performance inference pipelines for large language models on GPUs. The role focuses on profiling and tuning model execution, memory management, speculative decoding, context caching, quantization, and benchmarking systems. Candidates need a relevant bachelor's, master's, or higher degree, at least five years of software development experience, strong Python and software engineering skills, and experience with PyTorch and HuggingFace. The position is hybrid.
