900-21030-2220-000 NVIDIA TESLA C2070 GRAPHICS CARD
Based on the new NVIDIA CUDA GPU architecture codenamed "Fermi", the NVIDIA Tesla C2070 900-21030-2220-000 graphics card was built from the ground up for high-performance computing. It supports "must have" features for tech computing, enterprise computing, and virtual reality gaming.
As a dual-slot graphics card, the NVIDIA Tesla C2070 draws power from a 1x 8-pin power connector, with a rated power draw of 238W max. Display outputs include: 1x DVI. The Tesla C2070 GPU is connected to the rest of the system using a PCI-Express 2.0 x16 interface. The card is 248mm long, and features a dual-slot cooling solution.
It features 448 shaders, 56 texture mapping units, and 48 ROPs. NVIDIA has paired 6GB of GDDR5 memory with the Tesla C2070, which communicates using a 384-bit memory interface. The GPU operates at a frequency of 574MHz; Memory operates at 747MHz (3Gbps effective). The card has 4 GPCs each capable of 8 pixels per hour. This limits the full GPU to 32 pixels per hour and because of that it cannot feed all 48 ROPs when they all require data at the same time.