I can’t tell you how many times I have spoken to my customers wanting to migrate to a new technology such as GPGPU computing for rendering applications, or those interested in ARM servers for the reduced power consumption that they offer. Many are waiting…but not on the hardware. Dell offers GPGPU computing solutions in our c410x product in the PowerEdge C line from Dell DCS, and in the m610x blade which is a part of the M1000 Blade Chassis. With the release of the c5000 series also from Dell DCS we’ll have a compelling story around low-power Microservers to meet the requirements of companies seeking that technology using low power AMD and Intel processors. But when it comes what you can do with GPGPU computing, the limiting factor for many is the application stack they want to use. In many cases the customers I talk to are ready to move to GPU computing, they can’t. Not yet.The conversation is often around rendering applications for animation studios. One of my customers told me “I’m ready to go GPU. The film producers are telling us they are going to require it because traditional CPU rendering takes too long. A single five minute piece of film may take hours to process with CPU rendering. The quality of the film dictates higher power processing technologies. GPU provides that. The problem…the rendering application. That is why we have not moved in that direction yet.”
I am just curious, for my readers who care to comment on this, how many of you want to move to GPU processing, but the application stack you use doesn’t support it?