Nvidia Founder, President and CEO Jen-Hsun Huang delivers a keynote address at CES 2017 at The Venetian Las Vegas on January 4, 2017 in Las Vegas, Nevada. (Photo : Getty Images/Ethan Miller)
While mainstream PC gamers are waiting for the GTX 1080 Ti, NVIDIA has launched their new line of Quadro video cards. The company revealed the new family of professional workstation GPUs at the recent SolidWorks World conference in Los Angeles.
NVIDIA's new Quadro GP100 GPU is aimed at professionals who work in the computer-aided design industry. The devices are also used for fields that require fast and powerful computations such as medical and technical simulations.
One of the new capabilities of the cards is better handling of virtual reality. Besides gaming, there are also other industries that use VR for their businesses. One example is IKEA's VR experience for customers that want to try out a product in VR before actually buying it.
The new NVIDIA Quadro GP100 has 16GB of HBM and 20 TFLOPs of 16-bit floating point precision computing power, Digital Trends has learned. Other Pascal-based Quadro GPUs include the P400, P600, P1000, P2000 and the P4000. They will be replacing the previous Maxwell and Kepler GPU-powered cards.
Since the new GP100 card has half the memory of the previous P6000 GPU, NVIDIA has introduced a new NVLink connector. The said connector allows more computational power by linking multiple GPUs and even supported CPUs.
NVIDIA's NVLink connector on the new Quadro GP100 card is located where the SLI should have been, Ars Technica reported. The card has 3584 CUDA cores with 224 texture units.
In terms of memory bandwidth, the GP100 also beats the P6000 with 720GB/s compared to 432GB/s, respectively. It is also more power efficient with a TDP of 235W compared to 250W on the older model.
NVIDIA is expected to launch the Quadro GP100 in March. The price tag has not been announced but professionals should prepare at least $6225 considering that that was the cost of the P6000.
Learn what Quadro cards can do:
