NVIDIA is developing processor and system architectures that accelerate machine learning, automotive and high performance computing (HPC) applications. We are seeking a strong candidate to do performance analysis and kernels development for NVIDIA's new architectures. Your work will play a critical role in shaping the future of deep learning hardware and software, ensuring optimal performance for next-generation AI applications. This position offers the opportunity to make a meaningful impact in a fast-moving, technology focused company.What you'll be doing:Design, develop, and optimize major layers in LLM (e.g attention, GEMM, inter-GPU communication) for NVIDIA's new architectures.Implement and fine-tune kernels to achieve optimal performance on NVIDIA GPUs.Conduct in-depth performance analysis of GPU kernels, including Attention and other critical operations.Identify bottlenecks, optimize resource utilization, and improve throughput, and power efficiencyCreate and maintain workloads and micro-benchmark suites to evaluate kernel performance across various hardware and software configurations.Generate performance projections, comparisons, and detailed analysis reports for internal and external stakeholders.Collaborate with architecture, software, and product teams to guide the development of next-generation deep learning hardware and software.What we need to see:MS or PhD in relevant discipline (CS, EE, Math)3+ years of industry experience in GPU programming or performance optimization for DL applications.Demonstrated experience in analyzing and improving the performance of GPU kernels, with measurable results (e.g. performance improvements, efficiency gains).Strong programming skills in C, C++, Perl, or PythonStrong background in computer architectureExcellent communication skills, both written and verbal.Strong organizational and time management abilities, with the ability to prioritize tasks effectively.Ways to stand out from the crowd:LLM FMHA or GEMM related development or optimization experience will be a plusExpertise in CUDA programming for GPU acceleration will be a plus.Expertise in GPU/CPU Core or MemSys architecture modeling will be a plus.#deeplearning
