Principal Architect, AI Networking

Nvidia

Confirmed live yesterday High trust

Remote

Quick summary

Work type: Remote
Location: Santa Clara, CAAustin, TX
Salary: $272,000–$431,250 / yr
Posted: 141 days ago
Freshness: Confirmed live yesterday
Nearby: 99+ roles within 25 mi

Market check

Salary context

Above market

How this pay compares to similar roles

Similar $209k

This role $352k

$118k most similar roles pay here $465k

This role pays more than 98% of similar roles. Most pay $171,850–$246,725 — the shaded band above. At the midpoint, this role pays about $352k versus about $209k for comparable roles.

Based on 239 similar postings.

Employer

About Nvidia

Nvidia is a leading designer of graphics processing units (GPUs) and system-on-chip units, powering gaming, professional visualization, data centers, and artificial intelligence workloads. Industry: Semiconductors & AI Computing

Nvidia currently has 896 open roles on FindRole.

Listed pay typically runs $184,000–$287,500 across 876 roles with salary data.

Most-posted roles

View all roles at Nvidia

At a glance

TL;DR · Principal Architect, AI Networking

Role Posting Log in to save

As a Principal Architect, AI Networking within the Networking Systems & Software Architecture group, you will lead the research agenda and architectural direction for how systems communicate at scale across GPUs, DPUs, NICs, and heterogeneous storage. You will define project scopes from scratch, conduct original research on RDMA, NVLink, and GPUDirect, and prototype next-generation networking solutions while driving hardware-software co-optimization. Your daily work involves investigating bottlenecks in communication runtimes for large-scale AI workloads like KV cache transfer and model parallelism, while integrating capabilities into serving stacks such as vLLM, SGLang, and TensorRT-LLM. You will utilize C, C++, Rust, and Python to solve infrastructure problems involving InfiniBand, RoCE, NCCL, UCX, and MPI. The role addresses critical data movement challenges between nodes and storage to support high-performance distributed AI training and inference.

What does a Architect earn?

Median $235750 from 31 postings across 17 companies.

See salary data

Skills

RDMA NVLink GPUDirect InfiniBand RoCE NCCL UCX MPI NVSHMEM CUDA C C++ Rust Python vLLM SGLang TensorRT-LLM Distributed Training Computer Architecture

What you'll do

Define the long-term technical vision for distributed AI communication systems across GPUs, storage, and nodes.
Conduct original research and prototype next-generation networking solutions using RDMA, NVLink, and GPUDirect.
Drive hardware-software co-optimization involving GPUs, DPUs, NICs, and network switches.
Identify and resolve fundamental bottlenecks in communication runtimes for large-scale AI workloads like KV cache transfer.
Integrate networking capabilities into production AI serving stacks such as vLLM, SGLang, and TensorRT-LLM.
Translate research breakthroughs into production-grade software that ships to the industry.
Publish original findings and represent NVIDIA in industry forums and standards bodies.
Mentor senior engineers across the organization on technical architecture and research goals.

What we're looking for

15+ years of experience in systems software and/or networking.
Expertise in high-performance networking including InfiniBand, RoCE, RDMA, and NVLink.
Experience with communication libraries such as NCCL, UCX, MPI, or NVSHMEM.
MS, PhD, or equivalent experience in Computer Science, Computer Engineering, Electrical Engineering, or a related field.
Deep understanding of computer architecture, memory hierarchies, DMA engines, and OS-level networking.
Understanding of ML systems concepts including transformer architectures, KV cache mechanics, and model parallelism.
Proficiency in programming languages including C, C++, Rust, and Python.
Experience with CUDA programming and NVIDIA GPU architecture is preferred.

Save