Work involves researching, implementing, and validating model and algorithm improvements for video/world generation (human appearance, motion, interaction), improving training/inference efficiency, defining training objectives and benchmarks, and translating research into production-grade checkpoints and integrations. Requires a PhD or equivalent, 8+ years applied research/industry experience, and strong hands-on skills in Python, PyTorch, C++, CUDA, and distributed training.

Role to analyze deep learning networks and develop compiler optimization algorithms for NVIDIA's Deep Learning Compiler used across data centers, devices, automotive, and robotics. Requires 3+ years relevant experience, strong C/C++ and Python skills, and experience with technologies such as MLIR, XLA, TVM, LLVM, CUDA/OpenCL and deep learning frameworks (e.g., PyTorch). Posted locations include Austin, Redmond, and Santa Clara (also remote options); stated base salary range USD 152,000–241,500.

Lead the technical vision, original research, and prototyping for distributed AI communication systems (GPU-to-GPU, GPU-to-storage, cross-node) using technologies such as RDMA, NVLink, GPUDirect and communication libraries (NCCL, UCX, MPI, NVSHMEM). Requires 15+ years in systems software/networking, strong GPU-accelerated systems and computer architecture knowledge, and proficiency in C/C++/Rust/Python.

NVIDIA is hiring a Senior Software Engineer to design and implement JAX core components, drive performance on NVIDIA products, and build tools for AI teams. Requires 6+ years relevant experience, strong C/C++ and Python skills, familiarity with ML frameworks (e.g., PyTorch, TensorFlow), and technical knowledge of CPU/GPU architectures, numeric libraries, and distributed training.

Lead technical roadmap and design of next-generation communication primitives and collective algorithms for scaling models to very large clusters. Work includes co-design with applications and silicon architects, optimizing for interconnects (NVLink, Spectrum-X, Quantum-X), and developing high-fidelity models/simulators. Requires 12+ years in HPC or distributed deep learning, deep familiarity with NCCL/UCX/UCC/NVSHMEM/MPI, RDMA/RoCE/InfiniBand, CUDA and GPU memory hierarchy, and experience with inference/serving engines.

Role requires 8+ years of hands-on deployment and support experience for hardware and software products, strong Linux administration, cluster provisioning, networking, scripting (Bash/Python/Ansible), experience with schedulers (SLURM/LSF/UGE), Kubernetes, benchmarking tools (HPL/NCCL/MLPerf), and up to 30% travel within the U.S. to embed with customers and partners.

Technical leadership role translating advances in ML interatomic potentials, generative modeling, and quantum chemistry into developer tools, APIs, and scientific software for NVIDIA ALCHEMI. Requires a PhD or equivalent, 8+ years building AI/ML software for chemistry or materials, and strong experience with GPU-accelerated computing and modern ML frameworks.

Join NVIDIA Research Singapore to research and design efficient generative AI and foundation models (LLMs, visual generation, robotics) optimized for NVIDIA GPUs. Responsibilities include publishing original research, collaborating with internal and external teams, mentoring interns, and transferring technology to product groups. Requires a PhD in CS or EE, ~4 years' experience, strong Python prototyping skills; C++ and CUDA and familiarity with PyTorch are a plus.

NVIDIA is hiring a Senior AI/HPC Engineer on the infrastructure Specialist team to deploy, manage and maintain AI/HPC GPU infrastructure in Linux environments. The role involves customer-facing architecture and implementation, cluster and network optimization (InfiniBand/Ethernet), scheduler and MPI/NCCL expertise, automation (Ansible/Salt/Puppet), Kubernetes, and documentation/knowledge transfer. Requires a related degree or equivalent and 5+ years of deployment/support experience.

NVIDIA is hiring a Senior Compiler Engineer for its Deep Learning & AI Compiler (DLC) team to analyze deep learning networks, develop compiler optimization algorithms, and collaborate with software framework and GPU architecture teams. The role requires 3+ years of relevant work or research experience, strong C/C++ and Python skills, and experience with compiler technologies such as MLIR, LLVM, XLA, or Triton; GPU/CUDA experience and deep learning framework knowledge (PyTorch, JAX) are listed as advantages.

NVIDIA seeks a Senior Deep Learning Performance Architect to develop GPU and system architectures, build power and performance models for AI inference, construct and evaluate deep learning algorithms, and collaborate across software, research, and product teams. Requires MS/PhD or equivalent, 5+ years' relevant experience, strong ML/math foundation, expert programming in C/C++/Python, and familiarity with GPU computing (CUDA) and HPC (MPI/OpenMP).

Work on infrastructure software and tools to improve efficiency and resiliency for large-scale pre-training, post-training, and inference. Requires 8+ years building software infrastructure for large-scale AI systems, proficiency in Python/C/C++, experience with observability stacks (e.g., ELK, Prometheus, Loki), distributed systems, debugging/root-cause analysis, and familiarity with RDMA/NCCL/ucx/libfabrics and deep learning frameworks (PyTorch, TensorFlow, JAX, Ray).

NVIDIA Research Singapore is hiring a Research Scientist to research, design and accelerate generative AI models (LLMs, foundation models, visual generation, robotics) with a focus on GPU-efficient computing. Responsibilities include publishing original research, collaborating across teams, mentoring interns, and transferring technology to product groups. Required: ~2 years' experience, degree in CS or EE, strong research record, Python prototyping skills; C++/CUDA and PyTorch knowledge are a plus.

Work with science and research institutes in Japan to promote and deploy NVIDIA GPU-accelerated solutions for machine learning and generative AI. Responsibilities include leading research collaborations, designing architectures and demos, creating and running proofs of concept, onboarding developers to NVIDIA AI platforms (NIMs, NeMo), and presenting technical topics. Requires 5+ years' experience in generative AI/deep learning, strong coding skills (Python, C/C++, Bash, Linux), and excellent Japanese and business-level English communication.

NVIDIA seeks a Solutions Architect (Data Science) in ANZ to enable GSIs and AI consultancies on generative AI and LLM solutions (Nemo/NIMs), build GenAI full-stack solutions, run PoCs/demos, and support GPU-accelerated deployments. Requires deep expertise in generative AI/LLMs, ML/DL frameworks, Python, GPU/parallel computing, container orchestration and 5+ years' related experience.

NVIDIA seeks a Senior Math Libraries Engineer to scope, design, and implement numerical dense linear algebra software for AI and HPC on GPUs. Role requires C++ experience (8+ years), expertise in kernel generation, JIT/composable library design, performance optimization, collaboration across teams, and technical leadership. Preferred experience includes CUDA, parallel programming models (MPI/OpenMP/OpenACC/pthreads), assembly for performance, Python scripting, and familiarity with ML/DL and GPU hardware.

Work on communication libraries (e.g., NCCL, NVSHMEM, UCX) and co-design hardware/software solutions to accelerate AI and HPC workloads. Responsibilities include identifying communication bottlenecks, designing and implementing new communication technologies, building proofs-of-concept and simulations at large GPU scales. Requires MS/PhD or equivalent, 5+ years experience, strong C/C++, parallel programming and Linux skills, and knowledge of networking and system architecture.

Senior Math Libraries Engineer role to design, implement, and optimize C++/CUDA systems for unstructured sparsity in deep learning and HPC on NVIDIA GPUs; enable integration with Python/PyTorch, evaluate performance on real applications, and collaborate with product and partners. Requires 6+ years of high-performance software development experience.

NVIDIA is hiring a Senior Deep Learning Software Engineer for inference to design, build, and optimize GPU-accelerated software for large-scale model serving. The role involves performance tuning of LLM, multimodal and generative AI models across NVIDIA accelerators, contributing to inference libraries (vLLM, SGLang, FlashInfer), and using tools like CUTLASS, Triton, NCCL and CUDA. Requires a Master's/PhD or equivalent experience and 6+ years of relevant software development experience with strong C/C++ skills; Python and production inference/deployment experience are advantageous.

NVIDIA seeks a Senior Software Engineer to design and implement JAX core components, drive performance on NVIDIA products, and collaborate with researchers and cross-functional teams. Requires 6+ years' experience, strong C/C++ and Python skills, knowledge of ML framework internals, CPU/GPU architectures, numeric libraries, and experience with distributed training and performance optimization.