NVIDIA Quadro P6000 packs 3840 cores

NVIDIA has rolled out a new and impressively powerful graphics card for the designers and scientist out there who need graphics with serious oomph. The card is called the NVIDIA Quadro P6000 and NVIDIA Pascal, what the company claims is its most powerful GPU architecture ever, fuels it. The top of the line in the family is the P6000 and it has 3840 CUDA parallel-processing cores.

The card also has 24GB of GDDR5X GPU memory and 12 TFLOPs of FP 32 performance. With so much power on hand, the P6000 is a very power hungry card consuming a maximum of 250W of power. The card needs a PCI Express 3.0 x 16 slot on the PC mainboard to operate. Display outputs include four DisplayPort 1.4 ports, a single DVI-D port, and an optional stereo port.

The card is large at 4.4" tall and 10.5" long meaning compact case users need not apply. NVIDIA is aiming the cards in this family at scientific visualization and compute problem resolution. The card will also excel at VR design. NVIDIA is showing off the P6000, and its smaller sibling the P5000 at Siggraph 2016. The P5000 has 2560 CUDA cores, 16GB of GDDR5X memory, 8.9 TFLOPs of FP 32 performance, 180W power consumption, and is the same size and has the same video outputs as the P6000. One thing that we don't know about either of these cards right now is price. I would assume these are very expensive.

"Often our artists are working with 50GB or higher datasets," says Steve May, CTO at PIXAR. "The ability to visualize scenes of this size interactively gives our artists the ability to make creative decisions more quickly. We're looking forward to testing the limits of Pascal and expect the benefits to our workflows to be huge."