AMD has announced availability of the AMD Accelerated Parallel Processing (APP) SDK v2.5. Addressing advances in CPU to GPU data throughput, this latest APP SDK is billed as a route for developers to take advantage of Accelerated Processing Units (APUs).
Unsurprisingly perhaps, AMD is releasing this updated SDK in the wake of announcements relating to its own A-Series APUs with their HD graphics capabilities and improved battery life.
Positioned for developers who want the power of GPU compute and parallel processing in heterogeneous computing platforms, the updated SDK is said to build on the OpenCL industry standard, which is intended to help developers reach a broader market by writing applications that run on a variety of device form factors.
The updated SDK reduces CPU-to-GPU bandwidth limitations; thereby achieving effective data transfer rates as high as 15GB per second, with the latest AMD A-Series APUs at least. These same APUs are also compliant with advanced user-focused functionalities such as gestural interfaces, multi-monitor support, 3D entertainment, and real-time image stabilization.
"Improving performance and programmability on AMD platforms is a key initiative for us and we work closely with developers to help us make the APU the best possible development platform," said Manju Hegde, corporate vice president, AMD Fusion Experience Program.
Other features of the AMD APP SDK v2.5 include enhancements to the OpenCL runtime designed to enable more efficient use of available GPU compute capabilities by reducing kernel launch times and PCIe transfer overheads.
In addition, the AMD APP SDK provides broad multi-GPU support on Windows platforms including support for APU plus discrete GPU that enables compute performance scaling across multiple GPUs, as well as AMD PowerXpress technology support for APU plus discrete GPU.