Senior AI Communication Architect for HPC — Equity Eligible

NVIDIA; Austin (TX); On-site; USD 272,000 - 431,250; Full time

Job summary

NVIDIA in Austin, Texas, is seeking a highly experienced professional to drive architecture leadership and develop next-generation communication libraries for their platforms. Candidates must possess a Ph.D. or M.S. in a relevant field with over 12 years of experience in high-performance computing. The role requires expertise in 3D parallelism and NVIDIA GPU architecture. A competitive salary range of $272,000 to $431,250 is offered, along with equity and benefits.

Qualifications

• 12+ years of industry experience in high-performance computing or distributed deep learning.
• Expert knowledge of NVIDIA GPU memory hierarchy (HBM3e/HBM4, L2 cache).
• Hands-on experience developing within Megatron-Core, DeepSpeed, or JAX/XLA.

Responsibilities

• Define the long‑term technical roadmap for communication libraries across platforms.
• Lead development of next‑generation communication primitives and collective algorithms.
• Collaborate with silicon architects to influence hardware specifications.

Skills

Deep understanding of 3D parallelism

Technical proficiency with NCCL, UCX, UCC, NVSHMEM, MPI

Advanced knowledge of TensorRT-LLM, vLLM, SGLang, NVIDIA Dynamo

Education

Ph.D. or M.S. in Computer Science or related field

Tools

CUDA programming models

RDMA

RoCE

low-level InfiniBand verbs

Job description

NVIDIA in Austin, Texas, is seeking a highly experienced professional to drive architecture leadership and develop next-generation communication libraries for their platforms. Candidates must possess a Ph.D. or M.S. in a relevant field with over 12 years of experience in high-performance computing. The role requires expertise in 3D parallelism and NVIDIA GPU architecture. A competitive salary range of $272,000 to $431,250 is offered, along with equity and benefits.