Unmatched Power. Unmatched Creative Freedom.
NVIDIA Quadro GP100—The World’s Most Powerful Workstation Graphics and Compute Card.
The NVIDIA Quadro GP100, powered by NVIDIA’s Pascal GPU architecture, is equipped with the most advanced visualization and simulation capabilities to meet the needs of the most demanding professional workflows. A single GP100 with 3580 FP32/1792 FP64 CUDA Cores has incredible horsepower to render photorealistic design concepts interactively, create extremely detailed 3D models, run intensive CAE simulations to validate designs, and evaluate design prototypes in immersive VR. You can even connect two GP100’s with NVIDIA NVLink technology to scale performance with an amazing 32GB of combined HBM2 ultra-high bandwidth memory.
Designs are becoming more complex. Media is becoming richer, higher fidelity, combining greater resolutions and complex visual effects. Scientific visualization and compute problems are larger than ever. NVIDIA Quadro Sync II compatibility enables the GP100 to build massive and sophisticated visualization, simulation and rendering environments. VR is changing all facets of entertainment, design, engineering, architecture and medicine. AI and deep learning are changing the way professionals are solving the most complex and challenging computing problems. Customers want to experience ideas, validate designs, test ideas, and simulate solutions and visualize the results as quickly as possible.
The GP100 is an excellent choice for developers who want to tap into the power of NVLink and Unified Memory to maximize their application performance. Additionally, its fast FP16 performance makes it an ideal desktop workstation solution for developers working on the latest Deep Learning/AI algorithms and frameworks.
|Host Interface||PCI Express 3.0 x16|
|Memory Technology||HBM 2 ECC|
|Memory Capacity||16 GB HBM2|
|Max Displays||4 Displays|
|Processor||NVIDIA Quadro GP100|
|Memory Bandwidth||Up to 717 GB/s|
|API Support||CUDA C, CUDA C++, DirectCompute 5.0, OpenCL, Java, Python, and Fortran|
|Graphics Resolution||Maximum DP 1.4 Resolution: HDR 7680 x 4320 at 30Hz (30-bit color) |
5K Display Support: HDR 5120 x 2880 at 60Hz (30-bit color)
4K Display Support: HDR 4096 x 2160 at 60Hz or 3840 x 2160 at 60Hz
Maximum DVI-D DL Resolution: 2560 x 1600 at 60Hz via 3rd party adapter
Maximum DVI-D SL Resolution: 1920 x 1200 at 60Hz via included adapter
|Streaming Processor Cores||3584 CUDA Parallel-Processing Cores|
|Peak Double Precision FP64 Performance||5.2 TFLOPS|
|Peak Single Precision FP32 Performance||10.3 TFLOPS|
|Peak Half Precision FP16 Performance||20.7 TFLOPS|
|Multi-GPU Scalability||NVLink (2-way) or SLI HB|
|NVIDIA CUDA™ Technology||Yes|
|Professional 3D Support||Yes, via included stereo connector bracket|
|Quadro Sync II Compatible||Yes (Frame Lock and Genlock)|
|NVIDIA GPU Direct Compatible||Yes|
|DisplayPort Output||x4 (with audio)|
|D-SUB Output||Via optional adapters|
|HDMI Output||Via optional adapters|
|Max Graphics Card Power (W)||235W|
|Minimum Recommended Power, Single Card (W)||500W|
|Thermal Solution||Ultra-quiet active fansink|
|Dimensions||4.40” H x 10.50” L|