Having conquered the gaming industry with an early lead in powerful computing, chip maker Nvidia Corp. is now targeting the enterprise for a repeat success story in the business world’s big emerging markets such as artificial intelligence and virtual reality.

Upgrading its newest Tesla servers, Nvidia today unveiled a rebranded Quadro product that virtualizes its graphics card designed for personal computer workstations. Dubbed Quadro vDWS, Nvidia looks to inject its cloud offerings with a big boost that drives processing power to edge devices such as lightweight laptops.

Before today’s announcement, Nvidia would ship out Quadro to clients. Now just months after launching its most powerful graphics chip yet, Nvidia seems particularly driven to rapidly roll out cloud-based enhancements for its budding enterprise hardware business, as competition from Advanced Micro Devices Inc. and Fujitsu Ltd. is on the rise.

“Customers really want Quadro in the datacenter, and to not have to ship it out to the end points. So we’re virtualizing it,” said John Fanelli, vice president of products at Nvidia.

 According to the company, the new Quadro vDWS runs both virtualized graphics and compute workloads on any virtual workstation or laptop linked to Nvidia’s Tesla-accelerated datacenters. With this degree of virtualization, Nvidia’s graphics processing units can also be partitioned amongst end users, allocating a percentage of the GPU to systems engineers, for instance, and another percentage to the design team.  The amped-up Quadro vDWS is immediately available across Nvidia’s distribution partners, including original equipment manufacturers and system vendors, including Cisco Systems Inc., Citrix Systems Inc., Dell Technologies Inc., Hewlett Packard Enterprise Co., IBM Corp., Lenovo Group Ltd. and VMware Inc.

Nvidia’s cloud push also seeks to answer growing demand for compute-intensive workflows in enterprise environments, spanning data analytics to virtual reality and deep learning initiatives. It all comes at a comfortable cost that could attract new users, according to Patrick Moorhead, president and principal analyst and Moor Insights & Strategy. 

“This new piece of software enables enterprises to run even higher performance workstation applications from a low performance client pc, tablet and even a phone,” Moorhead said. “This will allow enterprises to better share very expensive resources globally, increase security because the data is always inside of the data center and not on the edge, and enable increased use of resources in place as the companies couldn’t afford it before.”

With GPUs’ proven value in the datacenter, Nvidia considers the enterprise a goldmine of opportunity. Designed to scale workloads upwards of 20,000 users, Quadro vDWS sees emerging markets in oil and gas, media and entertainment, and manufacturing and design.

“We see this evolution, and for the last 10 years we’ve supported these emerging workloads with our quadro product,” Fanelli said speaking of Nvidia’s propensity to position its GPUs in the right place at the right time. Office trends towards photo realism, immersive design and collaboration with holodecks (think automobile design) are driving Nvidia’s product development, Fanelli explained.

Yet these areas of potential market growth are also some of Nvidia’s slower-growing businesses currently. Revenue from professional visualization for high-end rendering such as that used in VR rose just 10 percent, to $235 million, in Nvidia’s second-fiscal quarter, while automotive rose 19 percent to $142 million. That compares with Nvidia’s strongest sector of PC gaming, which jumped 52 percent, to $1.19 billion in revenue.

GPUs more prevalent than you think

Nevertheless, Nvidia sees plenty of room to grow, anticipating even more demand for GPU-driven architectures thanks to forced software updates to Windows 10. In fact, all 70 million units comprising today’s PC market are up for grabs as far as Nvidia is concerned, with GPU technology touching many more apps than one might expect.

“YouTube, Skype, WebX — people are collaborating and want a truly visual experience. Even browsers are incorporating more dynamic visuals and fewer drop down boxes. Apps like Photoshop are now mainstream,” said Fanelli, listing a string of familiar applications benefitting from GPU technology.

“The move from Windows 7 to Windows 10 requires 30 percent more computing power. All of these changes are making the GPU a requirement, not a ‘nice to have,’” Fanelli added.

Other upgrades to Nvidia’s lineup include a subset of services for Nvidia GRID, the company’s graphics virtualization platform. The new GRID vPC product supports 50 percent more instances of virtual desktop profiles on its newest Tesla P40 server, while also offering application-level monitoring and management. Of note is the enhanced insight for end-to-end management to provide virtual GPU visibility from administrative and end user perspectives. This is a major point of differentiation, according to Fanelli.

“If you’re a datacenter operator, you know what’s going on with servers, where bottlenecks are, and if a user will soon run out of storage. There are the same needs for virtualized environments,” Fanelli explained. “I can look at it at the host level or an individual server. I can also query that server to see how many users it’s configured for.”

Photo: Nvidia