Language :
Take your professional workflows to the next level with the NVIDIA RTX PRO 5000 Blackwell. Equipped with the latest NVIDIA architecture and 48GB of fast GDDR7 memory, this GPU delivers exceptional performance, graphics, and AI capabilities to supercharge next-gen design, data science, simulation, and AI workflows. With NVIDIA's AI-powered technologies, professionals can accelerate their workflows, bring complex projects to life, and achieve stunning results, all while enjoying seamless multitasking and improved collaboration.
Designed for professionals who demand the best, NVIDIA RTX PRO solutions deliver unparalleled performance, reliability, and support. Every GPU is rigorously tested for a wide range of design, engineering, and AI workflows and continually optimized through enterprise drivers. With extensive ISV certifications, robust IT management tools, and enterprise-grade support, RTX PRO workstations are the trusted choice for enterprise deployments and mission-critical work.
The NVIDIA Blackwell architecture combines breakthrough AI, ray tracing and neural graphics technology, with massive performance and memory increases to drive cutting-edge professional creative, design, and engineering workflows from anywhere. NVIDIA RTX PRO™ Blackwell Generation GPUs are the most advanced workstation GPUs offering high-performance, real-time ray tracing, AI-accelerated compute, and professional graphics rendering.
Blackwell Streaming Multiprocessor (SM) integrates next-gen CUDA cores and neural shaders to redefine AI-augmented graphics and compute performance. Featuring up to 1.4X higher FP32 throughput versus the previous generation, the Blackwell SM architecture scales to 12,800 CUDA cores per GPU for industrial-grade simulations, 3D modeling, and ray-traced rendering. Enhanced neural shaders embed AI directly into programmable pipelines, enabling hybrid workflows like RTX Mega Geometry and DLSS 4.0 multi-frame generation. With support for FP64, FP32, FP16, and BF16 precision modes, a larger L2 cache, and higher clock speeds, Blackwell SMs accelerate tensor operations, sparse neural networks, and photorealistic rendering for AECO, manufacturing, and media workflows. This unified design ensures seamless concurrency for ray tracing, AI inferencing, and compute tasks—empowering professionals to tackle next-gen AI-augmented workflows with unprecedented efficiency.
5th-Gen Tensor Cores accelerate deep learning matrix operations critical to neural network training and inference. These cores feature 3X higher throughput over the previous generation with the Fine-Grained Structured Sparsity feature enabled, optimizing performance in sparse neural networks. They support mixed-precision computing— including newly added FP4 support for massive throughput increases, along with TF32, BF16, FP16, FP8, and FP6 data types—enabling flexible AI model development. Independent floating-point and integer data paths streamline complex workloads by parallelizing computation and addressing tasks. Enhanced with DLSS 4 multi-frame generation, they enable real-time AI rendering and high-fidelity content creation. For developers, this translates to unprecedented capabilities to workstations with double the memory and 3X the AI processing power of the previous generation to accelerate local LLM prototyping, industrial-scale inferencing, and efficient deployment of next-gen AI models, all while maintaining backward compatibility with existing frameworks.
4th-Gen RT Cores enable true-to-life neural graphics by delivering up to 2X faster ray tracing performance over the prior generation, achieving photorealistic scene rendering with physically accurate shadows, reflections, and refractions. Create photoreal, physically accurate scenes and immersive 3D designs with neural graphics-based technologies, such as RTX Mega Geometry, enabling up to 100X more ray-traced triangles. These cores are engineered for industrial-grade precision—ideal for prototyping, immersive 3D environments, and M&E content creation—while dramatically reducing iteration cycles for unprecedented productivity gains. Working in concert with applications leveraging APIs such as NVIDIA OptiX, Microsoft DXR, and Vulkan ray tracing, systems based on RTX PRO GPUs will power truly interactive design workflows to provide immediate feedback for unprecedented levels of productivity.
NVIDIA Blackwell GPU architecture adds hardware encode and decode support for 4:2:2 chromasampled video, bringing huge performance benefits when working in this format, enabling professionals to spend less time waiting & juggling proxies, and more time delivering their next masterpiece. The additional color information retained by 4:2:2 over 4:2:0 can be especially helpful for HDR content, and for maintaining fine details like text or fine lines or workflows where the source will get color corrected over and over again like when color grading.
9th-gen NVIDIA NVENC engines significantly accelerate video encoding speed and improve quality for professional video applications. Support for 4:2:2 H.264 and HEVC encoding has been added, along with quality increases for AV1 and HEVC increasing by 5% BD-BR PSNR. There’s also a new AV1 Ultra High Quality (UHQ) mode that takes additional time and provides an extra 5% improvement for the best quality possible. NVENC can take on the most demanding 4K or 8K video encoding tasks to free up the graphics engine and the CPU for other operations - video encoding at 8K/60 is achievable for professional video editing. RTX PRO GPUs provide quality equivalent to medium preset of software-based encoders, but at 2X performance.
6th-gen NVIDIA NVDEC engines provide up to double H.264 decoding throughput compared to the previous generation and add support for 4:2:2 H.264 and HEVC decode. Professionals can benefit from high-quality video playback, accelerate video data ingestion, and use advanced AI-powered video editing features. The following video codecs are supported for hardware-accelerated decoding: MPEG-2, VC-1, H.264 (AVCHD), H.265 (HEVC), VP8, VP9, and AV1 video formats.
New and improved GDDR7 memory significantly boosts bandwidth and capacity, empowering your applications to run faster, and work with larger, more complex datasets. With 48 GB of GPU memory, tackle massive 3D and AI projects, explore large-scale VR environments, and drive larger multi-app workflows. GDDR7 memory delivers up to 2X memory bandwidth to accelerate ray tracing, rendering, and AI workloads. RTX PRO 5000 Blackwell Generation GPU provide industry-leading graphics memory footprints to address the largest datasets and models in latency-sensitive professional applications.
Native support for error correcting code (ECC) in GDDR7 memory allows for higher data reliability without any performance penalties or capacity overhead by enabling sideband ECC where a small region, separate from main memory, is used for ECC bits. ECC bits are used to correct single bit errors or flag double bit errors. RTX PRO GPUs enable meeting strict data integrity requirements for mission-critical applications with uncompromised computing accuracy and reliability for workstations.
Employing optimizations from the Max-Q family of technologies, the RTX PRO lineup includes form-factor optimized solutions that focus on maximizing efficiency in dense multi-GPU configurations and/or small form factor systems.
RTX PRO GPUs support PCI Express Gen 5, which provides up to 64 GB/sec in x16 connections (double the bandwidth of PCIe Gen 4), increasing data-transfer speeds from CPU memory for data-intensive tasks like AI and data science. Faster PCIe performance also accelerates GPU direct memory access (DMA) transfers, providing faster video data transfers from GPUDirect for video-enabled devices and faster input and output with GPUDirect Storage.
Offloads GPU context scheduling from the CPU to GPU, reducing bottlenecks/latency and freeing the CPU for critical tasks (e.g., data processing, I/O operations), improving responsiveness in complex workflows while maximizing GPU utilization for demanding creative, engineering, or scientific workloads. Optimizes concurrent AI and graphics tasks (e.g., AI denoising, real-time physics simulations), accelerating workflows in tools like Omniverse or AI-augmented design software for faster, seamless productivity.
RTX PRO 5000 Blackwell GPUs includes multi-instance GPU (MIG) virtualization and GPU partitioning capability. When configured for MIG operation, RTX PRO 5000 delivers up to 2 GPU instances. Robust fault isolation allows partitioning RTX PRO 5000 GPUs safely and securely.
Pixel-level preemption provides more granular control to better support time-sensitive tasks such as VR motion tracking.
Preemption at the instruction level provides finer-grain control over compute tasks to prevent long-running applications from either monopolizing system resources or timing out.
With DisplayPort 2.1b, professionals can achieve unparalleled visual clarity and performance, driving high-resolution displays at up to 8K at 240Hz and 16K at 60Hz. Increased bandwidth enables seamless multi-monitor setups, ideal for multitasking and collaboration, while HDR and higher color depth support ensures superior color accuracy for precision work, such as video editing, 3D design, and live broadcasting. Support up to four 4K displays @ 165Hz per card or up to two 8K displays @ 100Hz with DSC using DP or HDMI. Highest link rates require a DP80LL certified cable.
Gain unprecedented end-user control of the desktop experience for increased productivity in a single large display or multi-display environments, especially in the current age of large, widescreen displays.
Synchronize the display and image output of up to 32 displays2 from 8 GPUs (connected through two Sync boards) in a single system, reducing the number of machines needed to create an advanced video visualization environment. Or, expand beyond a single machine and synchronize hundreds of displays across multiple systems for large deployments using Sync. Experience the added benefit of Variable Refresh Rate (up to 1000Hz) when you pair an RTX PRO 5000 Blackwell GPU to the RTX PRO Sync add in card. VRR enables varying refresh rates on synchronized displays that help them adjust rapidly in areas of low latency.
Transparently scale the desktop and applications across up to 4 GPUs and 16 displays from a single workstation while delivering full performance and image quality. This enables resolutions up to 16K x 16K in Windows, up to 32K x 32K in Linux. Mosaic can be combined with RTX PRO Sync to synchronize displays in this configuration.
Each frame lock connector is designed with a self-locking retention mechanism to secure its connection with the frame lock cable to provide robust connectivity and maximum productivity.
Robust control of stereo effects through a dedicated connection to directly synchronize 3D stereo hardware to an NVIDIA RTX PRO graphics card.
NVIDIA App delivers a suite of productivity tools to your desktop workstation, including desktop recording in up to 8K, automatic alerts for the latest NVIDIA RTX Enterprise driver updates, and access gaming features. The application is available for download here.
Deep learning frameworks such as Caffe2, MXNet, CNTK, TensorFlow, and others deliver dramatically faster training times and higher multi-node training performance. GPU-accelerated libraries such as cuDNN, cuBLAS, and TensorRT delivers higher performance for both deep learning inference and High-Performance Computing (HPC) applications. Get started today using NVIDIA AI Workbench to run sample projects or create your own environments in just a few clicks, complete with one-click collaboration features for teams. When you’re ready to deploy, NVIDIA AI Enterprise offers a massive suite of enterprise-ready pretrained models, tools, and support – for deployments of any scale.
Natively execute standard programming languages like C/C++ and Fortran, and APIs such as OpenCL, OpenACC, and Direct Compute to accelerate techniques such as ray tracing, video and image processing, and computation fluid dynamics.
GPUDirect for Video speeds communication between the GPU and video I/O devices by avoiding unnecessary system memory copies and CPU overhead.
Maximize system uptime, seamlessly manage wide-scale deployments, and remotely control graphics and display settings for efficient operations.
1 Formerly branded NVIDIA Quadro Sync II, effective March 18, 2025.
2 Support requires Linux and NVIDIA RTX Enterprise Driver R387 or newer. NVIDIA SLI Certified System required for over 4 GPUs.
3 The NVIDIA app unifies the NVIDIA Control Panel, NVIDIA RTX Experience™ apps.
| Architecture | NVIDIA Blackwell Architecture |
| CUDA Parallel Processing Cores | 14,080 |
| NVIDIA Tensor Cores | 440 |
| NVIDIA RT Cores | 110 |
| AI Performance1 | 2223 AI TOPS2 |
| GPU Memory | 48 GB GDDR7 with ECC |
| Memory Interface | 384-bit |
| Memory Bandwidth | 1,344 GB/s |
| Max Power Consumption | 300W |
| Multi-Instance GPU | Up to 2x 48GB Up to 1x 96GB |
| Graphics Bus | PCI Express 5.0 x16 |
| Display Connectors | DP 2.1 (4) |
| Form Factor | 4.4” H x 10.5” L, FHFL Dual Slot |
| Product Weight | 1.219 kg |
| Thermal Solution | Blower Active Fan |
| NVIDIA® 3D Vision® and 3D Vision Pro | Support via 3-pin mini DIN |
| Frame Lock | Compatible (with NVIDIA RTX PRO Sync) |
| Power Connector | 1x PCIe CEM5 16-pin |
| NVENC | NVDEC | JPEG | 3x | 3x | 1x |
1 Peak rates are based on GPU boost clock.
2 Theoretical FP4 TOPS using the sparsity feature.
| NVIDIA RTX PRO™ 5000 Blackwell Generation | |||
|---|---|---|---|
| Language | Version | Description |  |
| Quick Guide | |||
| English | Null ( 2017/3/20 ) | Quadro Quick Installation Guide Total size: [ 997 KB ] |  |
| Multilanguage | V01 ( 2018/4/1 ) | Supporting Models :P400, P600, P620, P1000, P2000, P4000, P5000, P6000, GP100, K420, K620, K1200, K2200, M4000, M5000 Quick start guide for Quadro series Total size: [ 2348 KB ] | |
| (简体中文) | Null ( 2016/6/13 ) | Quadro快速入門指南 Total size: [ 1192 KB ] | |
| Tesla Data Sheet | |||
| English | Null ( 2016/10/20 ) | DGX-1 Total size: [ 1356 KB ] | |
| English | Null ( 2016/10/20 ) | Tesla P100 Total size: [ 947 KB ] | |
| English | Null ( 2016/10/20 ) | Tesla P40 Total size: [ 4317 KB ] | |
| English | Null ( 2016/10/20 ) | Tesla P4 Total size: [ 4814 KB ] | |
| English | Null ( 2016/10/20 ) | Tesla M40 24GB Total size: [ 5396 KB ] | |
| (繁體中文) | Null ( 2016/10/20 ) | Tesla P100規格書 Total size: [ 1854 KB ] | |
| (繁體中文) | Null ( 2016/10/20 ) | DGX-1深度學習系統規格書 Total size: [ 974 KB ] | |
| DM | |||
| English | Null ( 2016/6/13 ) | Quadro Full Series DM Total size: [ 3167 KB ] | |
| (繁體中文) | Null ( 2016/6/13 ) | Quadro全系列中文型錄 Total size: [ 18612 KB ] | |
| NVS Data Sheet | |||
| English | Null ( 2015/11/24 ) | NVS810 Total size: [ 1298 KB ] | |
| English | Null ( 2015/11/24 ) | NVS510 Total size: [ 1886 KB ] | |
| English | Null ( 2015/11/24 ) | NVS315 Total size: [ 1149 KB ] | |
| English | Null ( 2015/11/24 ) | NVS310 Total size: [ 1188 KB ] | |
| (繁體中文) | Null ( 2015/11/24 ) | NVS810規格書 Total size: [ 1008 KB ] | |
| (繁體中文) | Null ( 2015/11/24 ) | NVS510規格書 Total size: [ 1366 KB ] | |
| (繁體中文) | Null ( 2015/11/24 ) | NVS315規格書 Total size: [ 1359 KB ] | |
| (繁體中文) | Null ( 2015/11/24 ) | NVS310規格書 Total size: [ 1420 KB ] | |
| Tegra Data Sheet | |||
| (繁體中文) | Null ( 2016/6/13 ) | Jetson TX1開發套件規格書 Total size: [ 9592 KB ] | |
| (繁體中文) | Null ( 2016/6/13 ) | Jetson TK1開發套件規格書 Total size: [ 9191 KB ] | |
| Quadro Data Sheet | |||
| English | Null ( 2017/3/20 ) | Quadro GP100 Total size: [ 1896 KB ] | |
| English | Null ( 2016/9/14 ) | Quadro P6000 Total size: [ 376 KB ] | |
| English | Null ( 2016/9/14 ) | Quadro P5000 Total size: [ 374 KB ] | |
| English | Null ( 2017/3/20 ) | Quadro P4000 Total size: [ 1545 KB ] | |
| English | Null ( 2017/3/20 ) | Quadro P2000 Total size: [ 1441 KB ] | |
| English | Null ( 2017/3/20 ) | Quadro P1000 Total size: [ 574 KB ] | |
| English | Null ( 2017/3/20 ) | Quadro P600 Total size: [ 603 KB ] | |
| English | Null ( 2017/3/20 ) | Quadro P400 Total size: [ 1454 KB ] | |
| English | Null ( 2016/6/13 ) | Quadro M6000 24GB Total size: [ 687 KB ] | |
| English | Null ( 2015/11/24 ) | Quadro M5000 Total size: [ 692 KB ] | |
| English | Null ( 2015/11/24 ) | Quadro M4000 Total size: [ 684 KB ] | |
| English | Null ( 2016/6/13 ) | Quadro M2000 Total size: [ 578 KB ] | |
| English | Null ( 2015/11/24 ) | Quadro K2200 Total size: [ 589 KB ] | |
| English | Null ( 2015/11/24 ) | Quadro K1200 Total size: [ 577 KB ] | |
| English | Null ( 2015/11/24 ) | Quadro K620 Total size: [ 595 KB ] | |
| English | Null ( 2016/10/21 ) | Quadro K420 2GB Total size: [ 601 KB ] | |
| (Eastern Language) | Null ( 2017/3/20 ) | Quadro GP100 規格書 Total size: [ 1916 KB ] | |
| (繁體中文) | Null ( 2016/9/14 ) | Quadro P6000規格書 Total size: [ 503 KB ] | |
| (繁體中文) | Null ( 2016/9/14 ) | Quadro P5000規格書 Total size: [ 457 KB ] | |
| (繁體中文) | Null ( 2017/3/20 ) | Quadro P4000 規格書 Total size: [ 1636 KB ] | |
| (繁體中文) | Null ( 2017/3/20 ) | Quadro P2000 規格書 Total size: [ 1481 KB ] | |
| (繁體中文) | Null ( 2017/3/20 ) | Quadro P1000 規格書 Total size: [ 1443 KB ] | |
| (繁體中文) | Null ( 2017/3/20 ) | Quadro P600 規格書 Total size: [ 1504 KB ] | |
| (繁體中文) | Null ( 2017/3/20 ) | Quadro P400規格書 Total size: [ 1504 KB ] | |
| (繁體中文) | Null ( 2016/6/13 ) | Quadro M6000 24GB規格書 Total size: [ 3151 KB ] | |
| (繁體中文) | Null ( 2015/11/24 ) | Quadro M5000規格書 Total size: [ 678 KB ] | |
| (繁體中文) | Null ( 2015/11/24 ) | Quadro M4000規格書 Total size: [ 639 KB ] | |
| (繁體中文) | Null ( 2016/6/13 ) | Quadro M2000規格書 Total size: [ 2749 KB ] | |
| (繁體中文) | Null ( 2016/6/27 ) | Quadro K2200規格書 Total size: [ 2741 KB ] | |
| (繁體中文) | Null ( 2015/11/24 ) | Quadro K1200規格書 Total size: [ 8284 KB ] | |
| (繁體中文) | Null ( 2016/6/27 ) | Quadro K620規格書 Total size: [ 2867 KB ] | |
| (Thai) | Null ( 2016/10/20 ) | Quadro P6000 Total size: [ 2905 KB ] | |
| (Thai) | Null ( 2016/10/20 ) | Quadro P5000 Total size: [ 5806 KB ] | |
| (Vietnamese) | Null ( 2016/10/20 ) | Quadro P6000 Total size: [ 3424 KB ] | |
| (Vietnamese) | Null ( 2016/10/20 ) | Quadro P5000 Total size: [ 3550 KB ] | |
Future
