Nvidia has announced a new virtualisation software suite that can turn the company's Tesla GPU-accelerated servers into powerful workstations for particular resource-hungry applications.
Quadro vDWS runs graphic and compute workloads in a virtualised environment, offering the ability to create increased productivity, unified graphic and compute workloads and significantly raised performance for Linux users. It is aimed at users working on 3D, photo-realistic rendering, virtual reality (VR) and deep-learning projects.
Nvidia recommends Pascal-based Tesla GPU accelerators, which when combined with the new vDWS suite, offer up to 24GB of GPU memory, to better handle complex modelling.
It also claims much improved performance compared to previous GPU architectures, access to CUDA and Open CL workflows to streamline design and computer-aided engineering and, as mentioned, a performance boost for Linux users by offloading the H.264 graphics work to the cloud.
"The enterprise is transforming. Workflows are evolving to incorporate AI, photo-realism, virtual reality, and greater collaboration among employees. The Quadro visualization platform is evolving with the enterprise to provide the performance required," said Bob Pette, vice president of Professional Visualization at NVIDIA.
He continued: "With Quadro vDWS on Tesla-powered servers, businesses can tackle larger datasets, power the most demanding applications and meet the need for greater mobility."
The news comes alongside improvements to the NVIDIA GRID virtual PC (vPC) line, which offers up to 24 instances of 1GB virtual desktop (double the previous offering), with greater vGPU insight via a range of dashboards including VMware vRealize Operations (bROps), Citrix Director and XenCenter.
The whole combination is intended to offer system managers a better service with more equilibrium, even if there's a resource hog somewhere in the chain.
Alongside all this comes the Tesla P6, a new blade server GPU that joins the P4, P40 and P100, offering 16GB of memory, which can be split down to 16x1GB machines.
At this years GTC conference in San Jose, Nvidia announced the Volta based GPU range, which it claims has the performance of 100 CPUs.
The software is available now at 100 server locations worldwide, including offerings from Cisco, Citrix, Dell, HP, Lenovo and VMWare. New Pascal-based feature and function follows on 1 September.
Wikileaks Vault 7 suspect Joshua Schulte fingered by FBI after re-using smartphone passwords on his PCs
Joshua Schulte indicted on 13 counts relating to Vault 7 leaks and trading in images of child abuse
Alexa for Hospitality will link with existing systems so guests can order room service and control the air con
Massive volcanic eruptions could have warmed Mars' surface sufficiently for oceans to form
Examination of fruit flies' brains generated more than one billion data points for scientists to analyse