NVIDIA jobs

NVIDIA is hiring a Senior Software Architect for the GPU Communications Libraries team to co-design next-gen data center platforms and scalable communications software. The role involves designing communication technologies for AI and HPC workloads, exploring hardware-software co-design, and modeling GPU cluster performance. Requires 5+ years of experience, strong C/C++ skills, and expertise in parallel programming (MPI, SHMEM) and communication runtimes. United States, remote.

NVIDIA seeks a Deep Learning Framework Communications Engineer to integrate advanced communication technologies into AI stacks like PyTorch and TRT-LLM. The role focuses on optimizing multi-GPU communication, scaling across clusters, and collaborating with AI frameworks. Requires 5+ years of software engineering, deep learning framework experience, and proficiency in Python, C++, and CUDA. Located in the United States, remote.

Senior Systems Software Engineer – Accelerated Kubernetes Performance and Scale role at NVIDIA DGX Cloud focused on scaling AI infrastructure and reducing cost per token. Responsibilities include end-to-end performance analysis across Kubernetes stack, architectural improvements for hyperscale clusters, container latency optimization, open-source contributions, and confidential container scaling. Requires broad systems/Kubernetes expertise, 8+ years distributed systems experience, strong systems performance understanding. Located in United States, full-time, hybrid work preferred. Base salary range not specified.

NVIDIA is seeking Senior ML Compiler Engineers to develop machine learning compilers and AI systems. The role focuses on creating innovative compiler technologies, building efficient attention runtimes, optimizing LLM inference kernels, and automating development flows. Requires 6+ years of ML systems experience, including compilers and LLM inference kernels, with strong Python and C/C++ skills and expertise in deep learning frameworks like PyTorch and JAX. Located in the United States, remote.

NVIDIA's DGXC Data Services team is building cloud-native software services for managing data across hybrid and multi-cloud infrastructures for GPU-based training and inference. The role involves designing data management services, providing dataset catalogs, managing metadata and lifecycle, and collaborating with research teams on storage innovations like GPU direct storage. Requires 5+ years of experience, BS in Computer Science (or equivalent), experience with Kubernetes and cloud providers (AWS/GCP/Azure), and programming in Go, Python, C/C++, or Java. Based in the United States, remote.