AMD Says That It Will Become Easier for Applications to Use More GPU Power

GPUs already play a central role in the operation of visually-intensive applications, especially games, but they can also handle parallel processing tasks in place of the CPU, a use that, AMD says, should grow as architectures evolve.

At the moment, both AMD and NVIDIA have their respective GPU computing modules which, instead of rendering graphics, are primarily meant to run parallel processing tasks,

In other words, they can do the job that would normally be run by the CPU, at least when the software itself supports parallel processing.

Unfortunately, software engineers often have a hard time actually making software capable of leveraging this capability because they have to assign the right hardware to perform specific tasks (they have to decide what tasks the x86 CPU does and which ones the GPU takes over).

The fortunate development regarding this is that both Intel and Advanced Micro Devices are now making central processors with built-in graphics.

This architecture evolution is expected to make parallel software development much easier than before, although actual progress might take years to accomplish.

"Although OpenCL can make developing applications that leverage the power of the multi-core CPU or the GPU much easier, both OpenCL and other current programming models continue to maintain a very clear distinction between the GPU and the CPU,” said Lee Howes, an MTS programming models engineer at AMD.

“This distinction limits the range of algorithms that can be cleanly and easily implemented. Over time, as we continue to integrate the CPU and the GPU, we expect the range of algorithms that can efficiently be mapped to an APU to increase, and for it to become easier to develop applications that leverage both capabilities without the developer needing the detailed architectural knowledge that is currently necessary."

The other problem that has to be worked around is the latency and bandwidth limitations when offloading tasks from the CPU to the GPU.

Still, AMD has, all in all, a very optimistic view of the chances for GPUs to become more prevalent in processing.