The NVIDIA Quadro RTX 8000 graphics card is based on NVIDIA’s Turing architecture. The card offers ECC support, which is a required feature in some industries for demanding professional applications. The Quadro 8000 also supports 48GB of memory, but it can be combined with another card using NVLink for a total of 96GB of GDDR6 memory to tackle the largest rendering, AI, VR, or virtualization workloads.
The NVIDIA Quadro RTX 8000 is equipped with 4608 NVIDIA cuda cores and 576 Tensor cores. It also supports the industry’s first implementation of the new VirtualLink 3 port to connect high-resolution VR head-mounted displays.
Powered by Nvidia’s Turing architecture, the Quadro RTX 8000 offers 4608 CUDA cores, 576 Tensor cores, and 72 RT cores to render models and scenes with high-quality shadows, reflections, and refractions. Based on a 12nm process with about 18,600 million transistors, the GPU also features 4608 shading units, 288 texture mapping units and 96 ROPs, making it a breeze to connect up to four high-resolution monitors.
Compared to the Quadro RTX 6000’s 24GB memory, the RTX 8000 doubles that to 48GB GDDR6 ECC memory and has the ability to be combined with another 8000 card for a total of 96GB, which uses 384-bit bus with up to 672GB/s of bandwidth. This is possible with the NVIDIA NVLink HB Bridge offering 100GB/s bidirectional bandwidth to deliver the latest capabilities for ray tracing, deep learning, and advanced shading.
The Quadro RTX 8000 keeps cool with an active fansink which has a blower design. The blower-style is usually preferred in higher density server and workstation applications, but temperatures will vary depending on the cooling configurations. Unlike the Titan RTX, which includes a backplate, the Quadro RTX 8000 does not have this option. A single 6-pin and a single 8-pin power connection are located at the end of the graphics card where the blower cooler is located, keeping the cabling mess-free for front to back airflow designs.
The NVIDIA RTX 8000 is built for professionals across various industries in need of ray tracing, deep learning, and visual computing workloads. Featuring 4608 CUDA cores, this GPU can render projects to accurately depict real-life models and scenes. The Turing architecture and support for two RTX 8000’s with up to 96GB of GDDR6 ECC memory combined, can provide the performance for any demand.
If you know what you want but can't find the exact configuration you're looking for, have one of our knowledgeable sales staff contact you. Give us a list of the components you would like to incorporate into the system, and the quantities, if more than one. We will get back to you immediately with an official quote.