Principal Architect, AI Networking
Core
Leading the research agenda and architectural direction for NVIDIA's AI systems communication at scale, defining project scope from scratch, and translating research breakthroughs into production-grade software.
Role type
Principal Architect, AI Networking
Builds
Systems-level software for data movement between GPUs, nodes, and storage; production-grade networking solutions.
Domain
AI Infrastructure, High-Performance Networking, GPU Systems
Deliverable
production ML models
Required skills
High-performance networking (InfiniBand, RoCE, RDMA, NVLink), Communication libraries (NIXL, NCCL, UCX, MPI, NVSHMEM), GPU accelerated systems, Computer architecture, Memory hierarchies, DMA engines, OS-level networking, ML systems concepts (transformer architectures, KV cache, model parallelism), C/C++/Rust/Python programming
Preferred skills
ML inference frameworks (vLLM, SGLang, TensorRT-LLM), CUDA programming, NVIDIA GPU architecture expertise, Product strategy influence, Open-source contributions
Technologies
InfiniBand, RoCE, RDMA, NVLink, NIXL, NCCL, UCX, MPI, NVSHMEM, vLLM, SGLang, TensorRT-LLM, CUDA, C, C++, Rust, Python
Responsibilities
Setting long-term technical vision for distributed AI communication systems; Conducting original research and prototyping next-generation networking solutions; Driving hardware-software co-optimization with GPU, DPU, NIC, and network switch; Integrating networking capabilities into AI serving stacks; Publishing findings and representing NVIDIA in industry forums; Mentoring senior engineers
Seniority
Principal, strategy & mentorship
