NVIDIA has announced the new NVIDIA Quadro Virtual Data Center Workstation Software (vDWS). The virtualization software features enable the utilization of NVIDIA Tesla GPU-accelerated servers as powerful workstations and provide the resources required to address enterprise-wide virtual workspace demands.

NVIDIA Quadro is utilized in hundreds of professional workstation applications and has long been recognized as the industry standard for professional enterprise virtualization. NVIDIA Quadro vDWS enables virtualized graphics and computing workloads to be processed on any virtual workstation or laptop provided by NVIDIA Tesla accelerated data centers. Quadro vDWS is available on over 120 systems from 33 system manufacturers, and reduces the cost of ownership by delivering high performance to multiple enterprise users with the same GPU.

Quadro vDWS enables collaboration and high levels of mobility among a workforce scattered across the globe. It also addresses the challenges of highly computationally intensive workflows. Computationally intensive workflows are becoming increasingly severe due to the exponentially growing size and complexity of data, driven by the emergence of new technologies such as 3D, photorealistic rendering, virtual reality, and deep learning. This phenomenon is commonly observed, particularly in the fields of engineering and science, which are representative areas where simulations of the design process are executed to accurately predict final outcomes.

When equipped with Tesla GPU accelerators based on the NVIDIA Pascal architecture, Quadro vDWS supports : processing of large-scale immersive models with up to 24GB of GPU memory ; graphics performance improved by up to 2x compared to existing NVIDIA GPU architectures ; integration of graphics and computation workloads ; and performance improvements for Linux users.

Bob Pette, Vice President of Professional Virtualization at NVIDIA, said, “Enterprises are undergoing a period of transformation, and workflows are evolving to integrate artificial intelligence, realism, virtual reality, and enhanced collaboration among employees.” He also stated, “The Quadro virtualization platform is evolving to provide the performance that enterprises require, and through Quadro vDWS provided on Tesla-equipped servers, enterprises can solve large-scale datasets, run performance-intensive applications, and meet the needs for enhanced mobility.”

With Quadro vDWS, NVIDIA customers like the ESI Group can support multiple engineers with a single GPU, while simultaneously remotely accessing powerful applications in real time and performing immersive product virtualization.

“ESI’s customers are looking to deploy virtual reality across their entire enterprise,” said Eric Kam, Product Marketing Manager for the Immersive Experiences Business Unit at ESI. “Virtual prototypes can present similar challenges to physical prototypes. They are relatively expensive and can be limited to specific locations due to hardware and software constraints. With NVIDIA Quadro vDWS and Tesla P40, ESI’s VRify application delivers the same power, performance, and user experience as a physical workstation.” “This enables users to easily manage data-intensive 3D assets, integrate complex geometric structures into virtual environments, and process real-time multi-body physics simulations, and potentially allow users to experience their designs in virtual reality environments through a browser anywhere without performance degradation,” he said.

To address the growing demand for graphics-accelerated VDI, NVIDIA announced performance improvements to its NVIDIA GRID vPC product.

Grid vPC is a new feature included in the 'NVIDIA Grid August 2017 release' and 'Pascal-based NVIDIA Tesla GPU accelerator' that improves user density and scalability, enhances vGPU insights, and improves the user experience by consistently and smoothly allocating vGPU resources with a new GPU resource scheduler.

Following the NVIDIA Tesla P4, P40, and P100, the newly introduced Tesla P6 is designed for blade servers, offering 16GB of memory and supporting up to 16 instances on 1GB virtual desktop profiles. Pascal-based Tesla accelerators provide IT departments with the resources needed to meet and scale enterprise-wide demands for graphics and computational virtualization.