System Software Architect, AI and GPU Networking
Core
Design and prototype features for NVIDIA GPU Networking (Dynamo, NIXL, UCX) to accelerate AI workloads, data movement, and model serving in AI data centers.
Role type
Senior System Software Architect (AI Networking)
Builds
GPU networking libraries and runtime systems for distributed AI and HPC
Domain
AI Infrastructure, High Performance Computing, Networking
Deliverable
production ML models
Required skills
System architecture, C++, Python, CUDA, algorithm design, performance profiling, virtualization, networking, storage
Preferred skills
Research track record, Deep Learning frameworks (NCCL, MPI), AI communication libraries, Inference/Training workload optimization
Technologies
NVIDIA Dynamo, NIXL, UCX, GPUnetIO, NCCL, MPI
Responsibilities
Enhance GPU Networking offerings for AI workloads; Design and prototype features for data movement and model serving; Identify and evaluate new technologies and partner relationships; Develop enhancements to communication libraries; Evaluate innovative features for runtime systems.
Seniority
Senior, hands-on IC
