NVIDIA is hiring a Senior Deep Learning Software Infrastructure Engineer to craft, scale, and harden training libraries and infrastructure for large-scale autonomous driving model training. The role requires designing orchestration and distributed training systems, optimizing data loaders and performance, and collaborating with researchers and platform teams. Candidates should have 12+ years building high-performance distributed systems and strong experience with PyTorch, large-scale training (DDP/FSDP), NCCL, datacenter networking, parallel filesystems, schedulers (Slurm/Kubernetes), and Python.

NVIDIA seeks a Senior Developer Technology Engineer (DevTech) to research and develop techniques to GPU-accelerate high-performance database and ETL applications. The role involves analyzing and optimizing data-intensive workloads on current GPU architectures, collaborating with internal hardware/software teams and external experts, and influencing next-generation designs. Required: MSc/PhD (or equivalent), 5+ years relevant experience, strong C/C++ skills, low-level parallel programming (CUDA, OpenACC, OpenMP, MPI, pthreads, TBB), and deep knowledge of CPU/GPU memory subsystems and database/ETL domains.

Lead and scale an engineering team building and optimizing open-source GPU-accelerated deep learning inference frameworks (vLLM, SGLang, FlashInfer). Responsibilities include driving roadmap and execution for Client AI inference frameworks, partnering with compiler/libraries/research teams, performance tuning and profiling of LLMs and multimodal models, and guiding best practices for CUDA, Triton, CUTLASS and multi-GPU communications (NCCL, NVSHMEM). Requires MS/PhD or equivalent, 6+ years software development experience (3+ years technical leadership), strong C/C++ skills, hands-on GPU programming and proven experience deploying/optimizing deep learning models in production. Salary range provided: 184,000–356,500 USD/year.

Role at NVIDIA to lead architecture and performance work on Dynamo, building integrations with vLLM, SGLang, and TensorRT-LLM; responsibilities include finding and removing bottlenecks across runtimes, kernels, networking, and orchestration, and developing inference optimizations such as scheduling, disaggregation, KV caching, and autoscaling. Requires 3+ years experience and strong skills in Python/Rust/C++ and ML inference techniques.

NVIDIA is seeking an engineer to research, prototype, and implement systems optimizations at the intersection of deep learning frameworks and low-level CUDA. Responsibilities include designing and optimizing custom CUDA kernels, architecting distributed cluster-scale systems, analyzing hardware-software interactions, developing profiling and runtime tools, and collaborating with researchers and compiler/driver teams. Requires a BS/MS/PhD or equivalent, 2+ years relevant experience, strong C++ and Python skills, familiarity with CUDA and GPU architectures, and experience with transformers and distributed ML.

Work on research-oriented projects at the intersection of CUDA and deep learning systems, including custom high-performance CUDA kernel development, profiling and optimizing generative AI models and transformers, and architecting distributed systems that scale from single GPUs to supercomputer clusters. Requires strong C++ and Python skills, deep learning background, GPU/CUDA experience, and 8+ years of relevant experience.

Join NVIDIA's Cloud Foundations Automation team to architect, develop, and operate secure, scalable cloud-native platform services and infrastructure automation for DGX Cloud and SuperPod. Requires 8+ years' experience and strong proficiency in Python and Go, Kubernetes, infrastructure automation (Terraform, Ansible), workflow orchestration (Temporal), distributed systems (databases, Redis, Kafka/NATS/SQS), observability (Prometheus/Grafana/OpenTelemetry), networking, and Linux. Salary ranges provided in the posting.

Work on system software and firmware for SoCs in O-RAN Radio Units and Distributed Units, focusing on power-management policies (DVFS, gating, thermal), performance analysis, benchmarking, and validation across pre- and post-silicon environments. Collaborate with hardware, platform, and cellular-software teams to influence SoC and platform architectures.

Join NVIDIA's Performance and R&D teams to profile and benchmark AI and HPC workloads on GPU and CPU clusters, analyze networking and system-level performance (RDMA, MPI, NCCL, RoCE), develop diagnostic and benchmarking tools, and collaborate with hardware, firmware, networking, and software teams to produce reproducible performance insights.

Join NVIDIA's System Control Firmware team to design, implement, and secure low-level C firmware for the Boot and Power Management Processor (BPMP) in Tegra SoCs. Responsibilities include debugging hardware/software boundary issues, writing unit and integration tests, collaborating with global hardware and software teams, and supporting active product programs. Preferred experience includes embedded firmware, SoC power/clock/thermal management, and familiarity with Rust, ISO 26262, and secure boot concepts.

NVIDIA is hiring a Senior Software Engineer to build and operate automation software for network design, deployment, and day-2 management of GPU cloud and SuperPod deployments. The role requires 12+ years of relevant experience, strong proficiency in Python and Go, experience with NetBox/Nautobot and network configuration systems, familiarity with Kubernetes (including EKS), and streaming telemetry protocols (gRPC/GNMI). Published base salary ranges: $200,000–$322,000 (Level 5) and $248,000–$391,000 (Level 6) per year.

NVIDIA is hiring a senior software engineer to work on the CUDA driver for Microsoft Windows. Role involves architecting and implementing new features, coordinating across teams, and writing maintainable, well-tested C/C++ code that interacts with kernel- and user-mode components and OS interfaces. Requires a BS/MS (or equivalent) and 8+ years of relevant experience, strong C/C++ skills, experience with OS threads/process/virtual memory, multithreaded debugging, and system-level architecture knowledge. Preferred: WDDM, Windows device drivers, kernel-mode development, and graphics API experience.

Join NVIDIA's NVHPC compilers & tools group to analyze and run HPC applications on servers and distributed-memory systems, identify optimization opportunities, assist customers with GPU acceleration, and drive compiler and application engineering efforts. Requires 8+ years programming experience, strong skills in Fortran/C/C++, OpenACC/OpenMP/MPI/CUDA, performance analysis and tuning, and strong communication skills.

Embed with internal customers to provide architectural guidance, build code and tooling, and drive GPU capacity and infrastructure efficiency across compute, storage, networking, and orchestration layers. Requires 12+ years building distributed systems and cloud infrastructure, production coding experience (Golang/Java/C/C++/Python/Rust), and expertise with Kubernetes, storage, networking, and deploying AI/ML workloads at scale.

NVIDIA is hiring a Senior Software Engineer (Resilience Engineering) for DGX Cloud to lead org-wide reliability strategy, establish and maintain SLOs, lead high-severity incident response, implement chaos engineering and failure-injection practices, and improve production code and tooling. The role asks for 8+ years of industry experience, deep experience running large-scale production systems, practical reliability experience, and proficiency in Go, Python, or similar languages. Listed salary range: USD 184,000–356,500 (base, annual across levels).

NVIDIA seeks a Senior Performance Engineer to design and execute performance studies, establish baselines, diagnose bottlenecks, and drive optimizations for DGX Cloud AI workloads. The role requires strong programming in C++ and Python, deep knowledge of operating systems, computer architecture, distributed systems, and experience with performance engineering, profiling, and benchmarking; experience with CUDA, GPU performance analysis, and deep learning frameworks (PyTorch, JAX/XLA) is listed as a differentiator.

Lead design and implementation of AI solutions (LLM-powered pipelines, agentic workflows) to accelerate post-silicon validation and automation across NVIDIA's silicon toolchain. Requires 5+ years building/deploying ML/AI systems, 2+ years owning applied AI/LLM workflows, strong Python and at least one static language, EE fundamentals, and hands-on silicon lab/debug experience. Salary ranges provided for Levels 3 and 4.

NVIDIA is hiring an Applied AI Engineer to architect and deploy LLM-powered validation pipelines and AI integrations across silicon design and verification workflows. The role requires 5+ years building and deploying ML/AI systems or data-intensive backend services, strong Python and at least one static language (C/C++/C#/Java/Scala), experience with silicon characterization and lab debug tools, and familiarity with modern deep learning frameworks and agent/orchestration tools. Locations listed are across the US (including Santa Clara, CA) and remote; posted salary ranges from 152,000 USD to 287,500 USD per year depending on level and location.

Role owns the end-to-end evaluation flywheel: strategy and architecture for turning driving and simulation data into curated golden datasets, measuring metrics (precision/recall), ensuring evaluation quality and release processes, building self-serve dataset and metric tooling, and partnering with engineers and AI model developers. Requires 12+ years building software with significant experience in autonomous vehicles, robotics, or large-scale ML systems.

Role to design and implement VLSI/EDA tooling and package design methodologies used at scale for NVIDIA multi-die AI/HPC platforms. Requires 5+ years building package design flows, strong Python/C++ fundamentals, AI/ML experience in engineering workflows, and solid electrical engineering knowledge (signal integrity, power delivery, parasitic extraction).