The Dream of the Thin Client 90’s is alive with NVIDIA GRID 2.0



Back in the day we called it Thin Client/Server Based Computing and the dream was ubiquitous access to desktops and app from anywhere, on any device. We worked really hard over the years and made great strides toward that goal. Every project since has been a laboratory for further improving compatibility, scalability, reliability and user experience. Great technologies from Citrix, Microsoft, VMWare, etc., have jet-fueled these efforts and the systems we deliver today are amazing. Yet for some uses cases, the user experience of remote desktops and applications didn’t quite reach 100% that of modern, high-powered local GPU based devices.

I have been an enthusiastic supporter of NVIDIA Grid capabilities as it overcame this very limitation for high end use cases. This has enabled scientists, engineers, designers, etc. to gain the same benefits mainstream businesses have enjoyed with Thin Client technologies for many years.



A laptop accessing an array of remote applications including automotive design, 3D modeling, image processing and video editing all  leveraging NVIDIA GPU


NVIDIA upped the ante today in their announcement from VMWorld 2015 of Grid version 2.0. They have expanded the serviceable use cases in important ways by increasing the user density/application performance, supporting a wider range of servers and blade systems, and, adding LINUX guest OS support. There is new software and new Maxwell based cards as well:


Tesla M60 Tesla M6
GPU Dual High-end Maxwell Single High-end Maxwell
CUDA Cores 4096 1536
Memory Size 16 GB GDDR5 8 GB GDDR5
H.264 1080p30 streams 36 18
GRID vGPU CCU 2 / 4 / 8 / 16 / 32 1 / 2 / 4 / 8 / 16
Form Factor PCIe 3.0 Dual Slot MXM
Power 240W / 300W (225W opt) 100W (75W opt)
Thermal active / passive bare board


These new products from NVIDIA expand both the existing high-end capabilities, but even more significantly, they provide a big experience boost for us humble business users. Why does that even matter, aren’t we just doing flat 2D applications? Well more and more these days, browsers and productivity software incorporate graphical and 3D processing, and it’s not just “eye candy”. Take for example the richness of the info you gain from your daily internet browsing and the time you spend on your latest PowerPoint presentation. Your experience is significantly enhanced by the addition of vGPU making your work much more productive (and enjoyable!). You get the all the benefits of virtualization and centrally hosted compute with the visual experience  you expect from a powerful local graphics processor.

The ability support to 128 users per system democratizes  this experience for the larger user sets we commonly deploy for corporate clients.When it comes to high-end users, i.e. the rocket scientist, we expect to leverage expensive hardware for smaller, more specialized users. Now awesome GPU goodness is available for all on commonly used rack servers and blade systems. So lets go out and re-start our XenApp/XenDesktop engines with vGPU and crank up the User Happiness factor. The Dream of the 90’s is Alive with NVIDIA GRID 2.0!