Senior Deep Learning Communication Architect

Description

The company has worked in computer graphics, PC gaming and accelerated computing for over 25 years and is now applying AI to a new phase of computing, where GPUs act as the brains of computers, robots and self-driving cars.

The software architecture group scales DNN models and training and inference frameworks to systems with hundreds of thousands of nodes.

The engineering teams are growing quickly and are looking for autonomous, creative engineers who enjoy technology.

Responsibilities

• Find and remove bottlenecks in data transfer and synchronization during distributed deep learning training and inference.
• Design and build communication algorithms and protocols for deep learning workloads that reduce overhead and latency.
• Work with hardware and software teams on systems that use high-speed interconnects such as NVLink, InfiniBand and SPC-X, along with communication libraries like MPI, NCCL, UCX, UCC and NVSHMEM.
• Research and assess new communication technologies and methods that improve the speed and scale of deep learning systems.
• Build proofs of concept, run experiments and create quantitative models to test and roll out new communication strategies.

Requirements

• A PhD, Master's or BS in computer science, electrical engineering, CSEE or a closely related field, or equivalent experience.
• 6+ years of work on building or scaling DNNs, DNN framework parallelism, or deep learning training and inference workloads.
• Hands-on experience evaluating and optimizing LLM training and inference performance for current models on leading-edge hardware.
• A deep understanding of data, pipeline, tensor and expert parallelism, and FSDP.
• Familiarity with emerging serving designs such as disaggregated serving, and with inference servers such as Dynamo and Triton.
• Experience writing code for one or more DNN training or inference frameworks, such as PyTorch, TensorRT-LLM, vLLM or SGLang.
• Strong programming skills in C++ and Python.
• Familiarity with GPU computing using CUDA and OpenCL, and with InfiniBand and RoCE networks.
• Ways to stand out (optional): prior contributions to one or more DNN training or inference frameworks in previous work.
• Ways to stand out (optional): deep understanding of, and contributions to, scaling LLMs on large-scale systems.

Conditions

Base salary depends on location, experience and other objective, job-related factors.

Level 4 base salary range: 184,000 USD to 287,500 USD.

Level 5 base salary range: 224,000 USD to 356,500 USD.

This is an existing vacancy, and applications are accepted at least until November 3, 2026.

The company uses AI tools in its recruiting processes.

The company is an equal opportunity employer and does not discriminate on protected characteristics, including in hiring and promotion.

Benefits

• Competitive salary and a generous benefits package.