Programming tools that harness the computing power of CPUs and graphics processors have been updated, bringing more parallel programming capabilities to the table.
Standards-setting firm Khronos Group has released OpenCL 2.0, which is a key development platform used to write applications in which processing is broken down over multiple processors and hardware inside systems. The group also released OpenGL 4.4, a graphics programming standard that takes advantage of the latest graphics hardware available in consoles, PCs and mobile devices.
OpenCL has grown in importance as graphics hardware and other co-processors are increasingly used to crunch complex math and science applications. Some of the world’s fastest computers combine CPUs and co-processors to speed up application processing, and Hewlett-Packard and Dell are offering servers and workstations loaded with graphics cards for customers that work on visual and CAD/CAM applications.
“It does significantly expand some of the new GPGPU compute function,” said Jim McGregor, principal analyst at Tirias Research.
McGregor referred to general-purpose graphics processing unit computing, in which processing is increasingly offloaded to graphics processors in systems.
But the effectiveness of OpenCL depends on programming and OS, which needs to support all the functions. OpenCL is backed by organizations such as Intel and Nvidia, which offer their own parallel programming tools to speed up processing of applications. Microsoft offers DirectX, its own parallel programming framework that is also used for game development and rendering.
Khronos also announced OpenGL 4.4, which allows “applications to incrementally use new features while portably accessing state-of-the-art graphics processing units (GPUs) across diverse operating systems and platforms,” the organization said in a release.
The new graphics specification also allows easy porting of applications across APIs (application programming interfaces), Khronos said.