# NVIDIA Developer

> Comprehensive developer portal for NVIDIA accelerated computing, AI, robotics, graphics, and simulation technologies.

## Getting Started

- [Platforms and Tools](https://developer.nvidia.com/platforms-and-tools.md): Explore NVIDIA platforms, SDKs, and developer tools by category
- [Developer Tools Catalog](https://developer.nvidia.com/developer-tools-catalog.md): Search the full catalog of NVIDIA developer tools, SDKs, and libraries
- [Open Source Catalog](https://developer.nvidia.com/open-source.md): Browse NVIDIA open source projects, libraries, and community contributions
- [Downloads](https://developer.nvidia.com/downloads.md): Download drivers, SDKs, toolkits, and firmware for NVIDIA hardware and software
- [Documentation](https://docs.nvidia.com/): Access technical documentation, API references, and programming guides across all NVIDIA products
- [Build API Catalog](https://build.nvidia.com/): Try and integrate the latest AI models, blueprints, and microservices with API endpoints
- [Developer Sandbox (Brev)](https://developer.nvidia.com/brev.md): Launch preconfigured cloud environments for prototyping and experimenting with NVIDIA tools
- [NGC Catalog](https://catalog.ngc.nvidia.com/): Browse GPU-optimized containers, pre-trained models, SDKs, and Helm charts for AI and HPC
- [AI Models](https://developer.nvidia.com/ai-models.md): Discover pre-trained models, containers, and resources for AI development on NGC
- [Topics](https://developer.nvidia.com/topics.md): Browse all developer topics across AI, simulation, graphics, HPC, and more

## Generative AI

- [AI Developer Resources](https://developer.nvidia.com/topics/ai.md): Overview of all NVIDIA AI tools, frameworks, and resources for developers
- [Generative AI](https://developer.nvidia.com/topics/ai/generative-ai.md): Overview of tools and models for generating text, image, audio, and video content
- [AI Inference](https://developer.nvidia.com/topics/ai/ai-inference.md): Overview of tools for deploying and optimizing AI inference models in production
- [Retrieval-Augmented Generation](https://developer.nvidia.com/topics/ai/retrieval-augmented-generation.md): Overview of RAG tools for grounding AI output with external knowledge sources
- [NeMo Customizer](https://developer.nvidia.com/nemo-customizer.md): Fine-tune and adapt large language models using supervised and parameter-efficient techniques
- [NeMo Evaluator](https://developer.nvidia.com/nemo-evaluator.md): Evaluate LLM quality with automated benchmarks, human preference metrics, and safety checks
- [NeMo Guardrails](https://developer.nvidia.com/nemo-guardrails.md): Add programmable safety, topic control, and content moderation to LLM-based applications
- [NeMo Agent Toolkit](https://developer.nvidia.com/nemo-agent-toolkit.md): Build multi-step agentic AI workflows with tool use, planning, and memory capabilities
- [NeMo Retriever](https://developer.nvidia.com/nemo-retriever.md): Deploy retrieval-augmented generation pipelines with GPU-accelerated embedding and reranking
- [NeMo Curator](https://developer.nvidia.com/nemo-curator.md): Curate, deduplicate, and filter large-scale training datasets for language model development
- [Nemotron](https://developer.nvidia.com/nemotron.md): Access open-weight LLMs with training recipes optimized for customization and deployment
- [Megatron Core](https://developer.nvidia.com/megatron-core.md): Train large language models at scale with GPU-optimized parallelism and mixed precision
- [TAO Toolkit](https://developer.nvidia.com/tao-toolkit.md): Customize pre-trained AI models with transfer learning for vision, speech, and language tasks
- [DALI](https://developer.nvidia.com/dali.md): Accelerate data loading and augmentation pipelines on GPUs for deep learning training
- [NIM](https://developer.nvidia.com/nim.md): Deploy optimized inference microservices for foundation models with a single API call
- [Dynamo](https://developer.nvidia.com/dynamo.md): Serve LLMs at scale with disaggregated inference, KV-aware routing, and SLA-based autoscaling
- [Dynamo/Triton Inference Server](https://developer.nvidia.com/dynamo-triton.md): Deploy multi-framework AI models in production with dynamic batching and model ensemble support
- [TensorRT](https://developer.nvidia.com/tensorrt.md): Optimize and deploy deep learning models for high-throughput, low-latency GPU inference
- [TensorRT-LLM](https://developer.nvidia.com/tensorrt-llm.md): Compile and optimize large language models for production GPU inference with quantization support
- [Riva](https://developer.nvidia.com/riva.md): Build speech AI applications with GPU-accelerated ASR, TTS, and neural machine translation
- [Maxine](https://developer.nvidia.com/maxine.md): Integrate AI-powered audio, video, and augmented reality effects into communication applications

## Accelerated Computing

- [CUDA Platform](https://developer.nvidia.com/cuda.md): NVIDIA's parallel computing platform for building GPU-accelerated applications
- [Data Science](https://developer.nvidia.com/topics/ai/data-science.md): Overview of GPU-accelerated tools for data processing, analytics, and machine learning
- [CUDA Toolkit](https://developer.nvidia.com/cuda/toolkit): Develop GPU-accelerated applications with compilers, libraries, and debugging tools
- [CUDA Python](https://developer.nvidia.com/cuda/python): Access CUDA runtime and driver APIs directly from Python for GPU programming
- [CUDA-X Data Science / RAPIDS](https://developer.nvidia.com/topics/ai/data-science/cuda-x-data-science-libraries.md): Accelerate data science workflows with GPU-powered analytics, ML, and ETL libraries
- [cuDF](https://developer.nvidia.com/topics/ai/data-science/cuda-x-data-science-libraries/cudf.md): Process DataFrames on GPUs with a pandas-compatible API for accelerated data manipulation
- [cuML](https://developer.nvidia.com/topics/ai/data-science/cuda-x-data-science-libraries/cuml.md): Train machine learning models on GPUs with scikit-learn-compatible algorithms
- [cuVS](https://developer.nvidia.com/cuvs.md): Perform GPU-accelerated vector search and nearest-neighbor retrieval for RAG and recommendation
- [NCCL](https://developer.nvidia.com/nccl.md): Coordinate multi-GPU and multi-node collective communication with topology-aware routing
- [NVSHMEM](https://developer.nvidia.com/nvshmem.md): Enable GPU-initiated one-sided communication across distributed memory for multi-GPU clusters
- [Nsight Developer Tools](https://developer.nvidia.com/tools-overview.md): Overview of profiling, debugging, and optimization tools for GPU-accelerated applications
- [Nsight Systems](https://developer.nvidia.com/nsight-systems.md): Profile system-wide CPU/GPU performance with timeline visualization and bottleneck analysis
- [Nsight Compute](https://developer.nvidia.com/nsight-compute.md): Analyze CUDA kernel performance with detailed hardware metrics and optimization guidance
- [Nsight Graphics](https://developer.nvidia.com/nsight-graphics.md): Debug and profile graphics applications across DirectX, Vulkan, and OpenGL APIs
- [CUDA-GDB](https://developer.nvidia.com/cuda-gdb.md): Debug CUDA GPU kernels and host code interactively with breakpoints and variable inspection
- [Compute Sanitizer](https://developer.nvidia.com/compute-sanitizer.md): Detect memory errors, race conditions, and synchronization bugs in CUDA applications

## CUDA-X Libraries

- [cuBLAS](https://developer.nvidia.com/cublas.md): Accelerate dense linear algebra with GPU-optimized BLAS routines for matrix operations
- [cuDNN](https://developer.nvidia.com/cudnn.md): Accelerate deep neural network training and inference with GPU-optimized primitives
- [cuFFT](https://developer.nvidia.com/cufft.md): Compute Fast Fourier Transforms on GPUs for signal processing and scientific workloads
- [cuSOLVER](https://developer.nvidia.com/cusolver.md): Solve dense and sparse linear systems with GPU-accelerated factorization and eigensolvers
- [cuSPARSE](https://developer.nvidia.com/cusparse.md): Perform sparse matrix operations on GPUs for scientific computing and graph analytics
- [cuRAND](https://developer.nvidia.com/curand.md): Generate high-quality random numbers on GPUs for Monte Carlo simulations and sampling
- [cuTENSOR](https://developer.nvidia.com/cutensor.md): Accelerate tensor contractions and element-wise operations for scientific and ML workloads
- [cuDSS](https://developer.nvidia.com/cudss.md): Solve large sparse linear systems with GPU-accelerated direct solver methods
- [nvCOMP](https://developer.nvidia.com/nvcomp.md): Compress and decompress data on GPUs with high-throughput batched algorithms
- [Thrust](https://developer.nvidia.com/thrust.md): Write portable parallel algorithms in C++ using an STL-like interface targeting CUDA GPUs
- [cuPyNumeric](https://developer.nvidia.com/cupynumeric.md): Run NumPy programs on GPUs and distributed systems without code changes
- [NVPL](https://developer.nvidia.com/nvpl.md): Access CPU-optimized math libraries for Arm-based NVIDIA Grace platforms
- [Nvmath-python](https://developer.nvidia.com/nvmath-python.md): Call CUDA math libraries from Python with a high-level pythonic API
- [cuEquivariance](https://developer.nvidia.com/cuequivariance.md): Accelerate equivariant neural networks with optimized CUDA kernels for geometric deep learning
- [cuLitho](https://developer.nvidia.com/culitho.md): Accelerate computational lithography for semiconductor manufacturing with GPU compute
- [Warp](https://developer.nvidia.com/warp-python.md): Write GPU-accelerated simulation and spatial computing kernels in Python
- [CUPTI](https://developer.nvidia.com/cupti.md): Instrument and trace CUDA applications programmatically for custom profiling tools

## Simulation and Physical AI

- [Design and Simulation](https://developer.nvidia.com/topics/design-and-simulation.md): Overview of developer resources for simulation, digital twins, and computer-aided engineering
- [Computer Aided Engineering](https://developer.nvidia.com/topics/cae.md): Overview of GPU-accelerated tools for CAE simulation and computational engineering
- [Omniverse](https://developer.nvidia.com/omniverse.md): Build and operate physically accurate 3D simulations and digital twins with OpenUSD and RTX
- [OpenUSD](https://developer.nvidia.com/openusd): Author, compose, and simulate 3D scenes using the Universal Scene Description framework
- [ACE](https://developer.nvidia.com/ace-for-games.md): Create AI-driven digital humans with speech, animation, and conversational intelligence
- [Newton Physics](https://developer.nvidia.com/newton-physics.md): Simulate rigid and soft body physics for robotics, gaming, and industrial applications
- [PhysX SDK](https://developer.nvidia.com/physx-sdk.md): Integrate real-time physics simulation for rigid bodies, fluids, cloth, and destruction effects
- [PhysicsNeMo](https://developer.nvidia.com/physicsnemo.md): Build and train physics-informed neural networks and neural operators for scientific simulation
- [Kaolin](https://developer.nvidia.com/kaolin.md): Accelerate 3D deep learning research with differentiable rendering and mesh operations
- [NanoVDB](https://developer.nvidia.com/nanovdb.md): Render sparse volumetric data on GPUs in real time with a lightweight VDB implementation
- [fVDB](https://developer.nvidia.com/fvdb.md): Train deep learning models on large-scale sparse 3D volumetric data
- [Warp](https://developer.nvidia.com/warp-python.md): Write GPU-accelerated simulation and spatial computing kernels in Python

## Robotics and Edge AI

- [Embedded Computing](https://developer.nvidia.com/embedded-computing.md): Overview of NVIDIA edge AI and embedded computing platforms for developers
- [Vision AI](https://developer.nvidia.com/computer-vision.md): Overview of tools for building applications that analyze images and videos with AI
- [Isaac](https://developer.nvidia.com/isaac.md): Develop and deploy AI-powered robots with end-to-end simulation, perception, and manipulation
- [Isaac ROS](https://developer.nvidia.com/isaac/ros.md): Add hardware-accelerated AI perception and navigation to ROS 2 robotics applications
- [Isaac Sim](https://developer.nvidia.com/isaac/sim.md): Simulate and test robots in physically accurate 3D environments with synthetic data generation
- [Isaac Lab](https://developer.nvidia.com/isaac/lab.md): Train robot policies with reinforcement learning and imitation learning in simulation
- [Isaac GR00T](https://developer.nvidia.com/isaac/gr00t.md): Develop general-purpose humanoid robot foundation models for dexterous manipulation
- [Jetson Platform](https://developer.nvidia.com/embedded/jetson-developer-kits.md): Build edge AI and robotics applications on compact, energy-efficient GPU modules
- [Jetson Modules](https://developer.nvidia.com/embedded/jetson-modules.md): Deploy edge AI on production-ready Jetson modules for commercial and industrial products
- [JetPack SDK](https://developer.nvidia.com/embedded/jetpack.md): Develop on Jetson with a complete BSP, CUDA toolkit, and AI libraries in one package
- [DeepStream SDK](https://developer.nvidia.com/deepstream-sdk.md): Build GPU-accelerated video analytics pipelines for multi-stream, multi-sensor AI at the edge
- [Holoscan SDK](https://developer.nvidia.com/holoscan-sdk.md): Process real-time sensor data with AI at the edge for medical devices and industrial systems
- [DGX Spark](https://developer.nvidia.com/topics/ai/dgx-spark): Develop and run AI workloads locally on a desktop-class NVIDIA Grace Blackwell system
- [Fleet Command](https://developer.nvidia.com/fleet-command.md): Deploy, manage, and update AI applications across distributed edge infrastructure
- [IGX Orin](https://developer.nvidia.com/igx-downloads.md): Build safety-certified edge AI applications for industrial and healthcare environments

## Autonomous Vehicles

- [DRIVE Platform](https://developer.nvidia.com/drive.md): Develop autonomous vehicle software with end-to-end simulation, perception, and planning tools
- [DRIVE OS](https://developer.nvidia.com/drive/os.md): Run safety-certified autonomous driving workloads on NVIDIA DRIVE hardware
- [DriveWorks SDK](https://developer.nvidia.com/drive/driveworks.md): Access sensor abstraction, calibration, and perception modules for autonomous driving pipelines
- [DRIVE AGX](https://developer.nvidia.com/drive/agx.md): Prototype autonomous vehicle applications on production-grade AI compute hardware
- [DRIVE Sim](https://developer.nvidia.com/drive/simulation.md): Test and validate autonomous driving software in physically accurate virtual environments
- [DRIVE Infrastructure](https://developer.nvidia.com/drive/infrastructure.md): Manage data pipelines and fleet operations for autonomous vehicle development at scale

## Graphics and Rendering

- [Ray Tracing](https://developer.nvidia.com/rtx/ray-tracing.md): Overview of NVIDIA ray tracing technologies, SDKs, and integration guides
- [Game Engines](https://developer.nvidia.com/game-engines.md): Integrate NVIDIA technologies into Unity, Unreal Engine, and other game engines
- [RTX Kit](https://developer.nvidia.com/rtx-kit.md): Integrate neural rendering and ray tracing technologies for photorealistic real-time graphics
- [DLSS](https://developer.nvidia.com/rtx/dlss.md): Boost frame rates and image quality with AI-powered super resolution and ray reconstruction
- [OptiX](https://developer.nvidia.com/rtx/ray-tracing/optix.md): Build GPU-accelerated ray tracing applications for rendering and scientific visualization
- [Reflex](https://developer.nvidia.com/performance-rendering-tools/reflex.md): Reduce system latency in competitive games with GPU-to-display pipeline optimization
- [Streamline](https://developer.nvidia.com/rtx/streamline.md): Integrate super resolution and latency reduction technologies via a single cross-vendor plugin
- [Vulkan](https://developer.nvidia.com/vulkan.md): Develop high-performance graphics and compute applications with the Vulkan GPU API
- [Extended Reality (XR)](https://developer.nvidia.com/xr.md): Build immersive AR, VR, and mixed reality experiences with NVIDIA XR technologies
- [AI Apps for RTX PCs](https://developer.nvidia.com/ai-apps-for-rtx-pcs.md): Develop and deploy AI applications that run locally on NVIDIA RTX Windows hardware
- [CloudXR SDK](https://developer.nvidia.com/cloudxr-sdk.md): Stream high-fidelity XR experiences from GPU-powered servers to lightweight client devices
- [VRWorks](https://developer.nvidia.com/vrworks.md): Build high-performance VR applications with GPU-accelerated rendering and display APIs
- [PhysX SDK](https://developer.nvidia.com/physx-sdk.md): Integrate real-time physics simulation for rigid bodies, fluids, cloth, and destruction effects
- [Video Codec SDK](https://developer.nvidia.com/video-codec-sdk.md): Encode, decode, and transcode H.264, H.265, and AV1 video using GPU hardware acceleration

## Video and Image Processing

- [Video and Audio Solutions](https://developer.nvidia.com/video-and-audio-solutions.md): Overview of NVIDIA video, audio, and broadcast processing technologies
- [Image Processing](https://developer.nvidia.com/image-processing.md): Overview of GPU-accelerated image processing and analysis tools
- [Metropolis](https://developer.nvidia.com/metropolis.md): Build and deploy intelligent video analytics and smart space applications at scale
- [DeepStream SDK](https://developer.nvidia.com/deepstream-sdk.md): Build GPU-accelerated video analytics pipelines for multi-stream, multi-sensor AI at the edge
- [CV-CUDA](https://developer.nvidia.com/cv-cuda.md): Accelerate computer vision pre- and post-processing pipelines on GPUs for AI inference
- [RTX Video SDK](https://developer.nvidia.com/rtx-video-sdk.md): Integrate AI-enhanced video upscaling, HDR, and processing into applications
- [nvImageCodec](https://developer.nvidia.com/nvimagecodec.md): Decode and encode images on GPUs with support for JPEG, JPEG2000, and other formats
- [nvJPEG](https://developer.nvidia.com/nvjpeg.md): Decode and encode JPEG images on GPUs for high-throughput batch image processing
- [nvTIFF](https://developer.nvidia.com/nvtiff.md): Decode and encode TIFF images on GPUs for scientific imaging and geospatial workloads
- [NPP](https://developer.nvidia.com/npp.md): Process images and signals with GPU-accelerated filtering, transforms, and color conversion
- [Optical Flow SDK](https://developer.nvidia.com/optical-flow-sdk.md): Estimate dense optical flow and motion vectors using dedicated GPU hardware engines

## Networking

- [Networking](https://developer.nvidia.com/networking.md): Overview of NVIDIA networking platforms for InfiniBand, Ethernet, and DPU development
- [DOCA](https://developer.nvidia.com/networking/doca.md): Develop data center services on NVIDIA BlueField DPUs for networking, storage, and security
- [InfiniBand](https://developer.nvidia.com/networking/infiniband-software.md): Build high-bandwidth, low-latency cluster interconnects for AI and HPC workloads
- [HPC-X](https://developer.nvidia.com/networking/hpc-x.md): Deploy optimized MPI and SHMEM communication libraries for InfiniBand and Ethernet clusters
- [Ethernet Switch SDK](https://developer.nvidia.com/networking/ethernet-switch-sdk.md): Program NVIDIA Spectrum switches with routing, ACL, and telemetry APIs
- [Rivermax](https://developer.nvidia.com/networking/rivermax.md): Stream media and sensor data over IP with hardware-accelerated SMPTE 2110 support
- [Magnum IO](https://developer.nvidia.com/magnum-io.md): Optimize I/O and data movement across GPUs, networks, and storage in multi-node systems
- [GPUDirect Storage](https://developer.nvidia.com/gpudirect-storage.md): Transfer data directly between storage and GPU memory bypassing the CPU for faster I/O
- [Aerial](https://developer.nvidia.com/industries/telecommunications/ai-aerial): Build software-defined 5G and 6G RAN infrastructure on GPU-accelerated platforms
- [Sionna](https://developer.nvidia.com/sionna.md): Simulate and research 6G link-level wireless communication systems on GPUs

## Cloud and Infrastructure

- [Cloud-Native Technologies](https://developer.nvidia.com/cloud-native.md): Deploy and manage GPU-accelerated applications in cloud and containerized environments
- [DGX Cloud](https://developer.nvidia.com/dgx-cloud.md): Develop and train AI models on a fully managed multi-node GPU cloud platform
- [DGX Cloud Serverless / NVCF](https://developer.nvidia.com/dgx-cloud/nvcf): Deploy AI models as serverless endpoints with auto-scaling GPU inference
- [DGX Cloud Benchmarking](https://developer.nvidia.com/dgx-cloud/benchmarking.md): Benchmark AI training and inference performance with standardized templates and dashboards
- [Morpheus Cybersecurity](https://developer.nvidia.com/morpheus-cybersecurity.md): Build GPU-accelerated cybersecurity analytics pipelines for real-time threat detection
- [FLARE Federated Learning](https://developer.nvidia.com/flare.md): Train AI models across distributed datasets without centralizing sensitive data
- [DCGM](https://developer.nvidia.com/dcgm.md): Monitor GPU health, diagnostics, and utilization across data center clusters
- [NVML](https://developer.nvidia.com/management-library-nvml.md): Query and control GPU state programmatically for monitoring and management tools
- [Grace CPU](https://developer.nvidia.com/grace-cpu.md): Develop for NVIDIA's Arm-based data center CPU optimized for AI and HPC workloads

## Healthcare and Life Sciences

- [Isaac for Healthcare](https://developer.nvidia.com/isaac/healthcare.md): Build AI-powered surgical, diagnostic, and medical automation robotics systems
- [Clara Guardian](https://developer.nvidia.com/clara-guardian.md): Deploy multimodal AI smart sensors for patient monitoring in healthcare facilities
- [Holoscan SDK](https://developer.nvidia.com/holoscan-sdk.md): Process real-time sensor data with AI at the edge for medical devices and industrial systems

## Quantum Computing

- [CUDA-Q](https://developer.nvidia.com/cuda-q.md): Program hybrid quantum-classical algorithms and simulate quantum circuits on GPUs
- [CUDA-QX](https://developer.nvidia.com/cuda-qx.md): Extend CUDA-Q with optimized libraries for quantum chemistry and error correction
- [cuQuantum](https://developer.nvidia.com/cuquantum-sdk.md): Simulate quantum circuits at scale using GPU-accelerated statevector and tensor network methods
- [cuPQC](https://developer.nvidia.com/cupqc.md): Implement GPU-accelerated post-quantum cryptography algorithms for security research

## High-Performance Computing

- [HPC](https://developer.nvidia.com/hpc.md): Overview of NVIDIA high-performance computing tools, compilers, and libraries
- [HPC SDK](https://developer.nvidia.com/hpc-sdk.md): Build GPU-accelerated HPC applications with compilers, math libraries, and communication tools

## Developer Industry Solutions

- [AECO](https://developer.nvidia.com/industries/aeco.md): Developer resources for architecture, engineering, construction, and operations
- [Consumer Internet](https://developer.nvidia.com/industries/consumer-internet.md): Developer resources for recommendation systems, search, and consumer AI applications
- [Energy](https://developer.nvidia.com/industries/energy.md): GPU-accelerated solutions for seismic processing, grid management, and energy analytics
- [Financial Services](https://developer.nvidia.com/industries/financial-services.md): Developer tools for trading, risk modeling, fraud detection, and financial AI
- [Game Development](https://developer.nvidia.com/industries/game-development.md): SDKs, engines, and tools for building GPU-accelerated games and interactive experiences
- [Healthcare](https://developer.nvidia.com/industries/healthcare.md): Developer resources for medical imaging, genomics, clinical AI, and healthcare analytics
- [Higher Education](https://developer.nvidia.com/higher-education-and-research.md): Academic programs, research tools, and curriculum for AI and accelerated computing
- [Media and Entertainment](https://developer.nvidia.com/industries/media-and-entertainment.md): Tools for content creation, real-time rendering, broadcast, and visual effects
- [Public Sector](https://developer.nvidia.com/industries/public-sector.md): Developer resources for government, defense, and public safety AI applications
- [Restaurants and QSR](https://developer.nvidia.com/industries/restaurants.md): Developer resources for AI-powered restaurant and quick-service operations
- [Retail and CPG](https://developer.nvidia.com/industries/retail-consumer-packaged-goods-cpg.md): AI solutions for store analytics, demand forecasting, and customer intelligence
- [Telecommunications](https://developer.nvidia.com/industries/telecommunications.md): Developer platforms for 5G/6G RAN, edge AI, and network automation

## Resources

- [Technical Blog](https://developer.nvidia.com/blog): Technical articles, tutorials, and announcements for NVIDIA developers
- [Training / DLI](https://www.nvidia.com/en-us/training/): Self-paced and instructor-led courses on AI, accelerated computing, and data science
- [Developer Program](https://developer.nvidia.com/developer-program.md): Access SDKs, early releases, and community resources with a free developer account
- [Developer Forums](https://forums.developer.nvidia.com/): Ask questions, share projects, and get technical support from the NVIDIA developer community
- [Developer Discord](https://discord.gg/nvidiadeveloper): Join the NVIDIA developer community for real-time discussion and technical support
- [NVIDIA On-Demand](https://www.nvidia.com/en-us/on-demand/): Watch recorded sessions from GTC and other NVIDIA technical events
- [For Startups / Inception](https://www.nvidia.com/en-us/startups/): Get technical resources, co-marketing support, and hardware credits for AI startups