Senior Software Architect - Deep Learning and HPC Communications

NVIDIA, the world leader in accelerated computing, is seeking a Senior Software Architect for their GPU Communications Libraries and Networking team. This role focuses on building crucial communication libraries like NCCL, NVSHMEM, and UCX for scaling Deep Learning and HPC. The position involves co-designing next-gen data center platforms and scalable communications software.

The role is central to NVIDIA's work in groundbreaking developments in Artificial Intelligence, High Performance Computing, and Visualization. You'll be working with systems that run at scales of up to tens of thousands of GPUs, connected with high-speed interconnects. Your work will directly impact end-to-end application performance and help break performance barriers in next-generation systems.

The ideal candidate should have strong expertise in C/C++ programming, parallel programming models, and deep understanding of system architecture. Experience with CUDA programming, NVIDIA GPUs, and knowledge of high-performance networks like InfiniBand, RoCE, and NVLink would be advantageous. The role offers competitive compensation, including a base salary range of $184,000 - $356,500 USD, plus equity and benefits.

This is an outstanding opportunity to advance the state-of-the-art in GPU communications and deliver platforms the world has never seen before. You'll be working in a diverse, multi-national environment, contributing to technologies that power everything from artificial intelligence to autonomous cars.