The VGX K2 will not be available until early 2013, but Nvidia released some initial specs of its first cloud computing GPU product.
The K2 will be the flagship board to enable smartphones, tablets and PCs to access substantially more GPU performance over a cloud infrastructure and expand the processing horsepower of their own local devices. The K2 is essentially two merged Quadro K5000 workstation graphics cards. It integrates two GPUs with 2x1536 active processing cores and 2x4GB GDDR5 memory. The board is rated at a power consumption of 225 watts and receives a passive cooling solution.
Pricing has not been announced yet, but don't expect to sell this product for about the same as a single Quadro K5000, which is currently available for about $1,750 and up.
Nvidia will also be offering an entry level VGX version, which uses four Kepler GPUs with 192 active cores each. The board will feature 16 GB GDDR5 memory and run at 130 watts while retaining the same dual-slot form factor of the K2.
“The VGX platform has been developed to bring rich, interactive graphics to all enterprise virtual desktop users,” said Jeff Brown, general manager of the Professional Solutions Group at Nvidia. “With VGX K2 in the data center, designers and engineers who create the core intellectual property for their companies can now access their IP from any device and still enjoy workstation-class performance.”
First to support Nvidia's VGX technology will be Citrix XenDesktop with HDX 3DPro. In combination with Citrix XenDesktop and Citrix XenApp, the VGX K2 board provides users full compatibility and performance of all their graphics and GPU computing-intensive applications.