AMD commends Intel and ARM for supporting OpenCL

Won’t let anyone forget it was first

WERE NOT SURE what’s going on over at AMD, but in a blog post on AMD’s official blog, AMD’s product marketing director, John Taylor, is commending both Intel and ARM for adding support for OpenCL to its products. It does, of course, also tell the tale of how AMD has been supporting OpenCL since 2009, but that’s a different story, right?

OpenCL has been a slow starter, mostly thanks to Nvidia and its proprietary CUDA solution for GPGPU which gained a lot of early attention. However, as OpenCL will run on both CPUs and GPUs, it’s a much more interesting standard for software developers to work with, as they aren’t getting tied into one company’s solutions. Of course there’s also Microsoft’s DirectCompute, but DirectCompute is a standard that’s been even slower to take off.

The Khronos group is in charge of the OpenCL standard as well as OpenGL and several other cross-platform standards. The selling point of OpenCL is that it’s an open, cross-platform standard and, as such, has far fewer limitations than more platform or hardware specific options. Still, it hasn’t proven to be immensely popular with software developers as yet, despite all major operating systems and now also hardware manufacturers offering support for the standard.

With Intel and ARM on board, things should start to progress at a faster pace. AMD is very keen to point out its contributions and wants to remind us that it’s latest generation of graphics cards, as well as its soon to launch Fusion family of APU’s will be the ideal hardware for OpenCL applications. It will be interesting to see how adaptation of the OpenCL standard will grow with all the key players in the industry now supporting it. ARM is said to add OpenCL to all of its future products and as such every smartphone by 2014 will support OpenCL.

There’s no doubt that parallel processing has a place in the market, as it can offer tremendous benefits depending on the task at hand. Anyone that’s tried rendering or transcoding video using GPU acceleration will know what kind of benefits that are on offer and if this can become a standard feature in all software, then we, the end users have all kinds of benefits to reap from it. Not only will we be able to save time when performing mundane tasks that can take hours to process today, but we might also be able to save money by buying the right combination of hardware which will be cheaper than simply just buying a faster processor to get more computational power.S|A

The following two tabs change content below.