Senior Software Engineer, AI Networking

Role responsibilities

Design and implement ML-based tools and combinatorial optimization techniques to optimize LLM training and inference at data center scale. Collaborate across hardware and software teams to develop performance analysis insights and establish targets for NVIDIA supercomputers.

Requirements

Requires a PhD or Master's degree with 4+ years of experience applying ML to computer architecture and system optimization. Proficiency in PyTorch/TensorFlow, C++, Python, and knowledge of NVIDIA GPUs and networking libraries like NCCL is essential.

Key skills

Machine Learning, Combinatorial Optimization, Reinforcement Learning, PyTorch, TensorFlow, C++, Python, CUDA, NCCL, Distributed Deep Learning, Computer Architecture, Networking Protocols, GNNs, Bayesian Optimization, Performance Analysis, Data Curation Pipelines

Keywords

AI Networking, LLM, Deep Learning, DSE, HPC, RoCE, RDMA, Kineto, GNNs, Transformers, Supercomputers, Distributed Systems, Collective Communication, Host Channel Adapters, Switches, GPU, CPU, Bash, System Architecture, Performance Prediction, Data Pipelines, Multi-objective Optimization, Agentic Techniques, NVIDIA