During its keynote at GPU Technology Conference in San Jose, California, Nvidia CEO Jen-Hsun Huang announced the Nvidia VGX platform, a way of delivering GPU computing performance of a PC or a workstation to any and every connected device.

With the implementation of Nvidia VGX boards that initially feature four GPUs, each with 192 CUDA cores and 4GB of frame buffer per GPU, IT departments are able to easily acces a GPU-accelerated desktop via any connected device but with ultra-low latency remote display capabilities and solve complex challenges of "BYOD" (bring your own device) in a work environment.

The Nvidia VGX key technology breakthroughs, according to Nvidia, include the aforementioned Nvidia VGX boards, Nvidia's VGX GPU hypervisor that is a software layer which enables virtualization of the GPU as well as the Nvidia User Selectable Machines (USMs), or simply a manageability option that allows a certain configuration in delivering graphics capabilities to individual clients on the network depending on their need.

According to Nvidia, the VGX platform enables up to 100 users to be served from a single server equipped with Nvidia VGX board, which is a decent improvement over standard VDI (virtual desktop infrastructure) solutions. It also addresses issues like latency, sluggish interaction an limited application support.

Nvidia's CEO pointed out that Kepler is a GPU that was designed for cloud computing making it a first virtualized GPU. This enabled Nvidia to push harder into cloud computing and allowing GPUs to be simultaneously shared by multiple users, delivering ultra-fast streaming display capabilities and basically "shifting cloud computing into a new gear", whether you are looking at enterprise or gaming implementation.