AMD today announced its involvement in the development of one of the first industry benchmark testing suites for OpenCL. Released by SiSoftware, the OpenCL GPGPU benchmark suite is part of SiSoftware Sandra 2010, the latest version of the award-winning utility first introduced in 1997. The benchmark suite includes remote analysis, benchmarking and diagnostic features for PCs, servers, mobile devices and networks, and can be used to test OpenCL performance on ATI Stream technology.
Heavy computational workloads have traditionally been processed on a CPU but the industry is shifting to a new computing paradigm that relies more on the GPU or a combination of GPU and CPU. OpenCL is the widely adopted industry standard for running parallel tasks on CPUs and GPUs using the same code. As the only hardware provider in the industry designing and delivering both high-performance CPU and GPU technologies, AMD is the only company providing a complete OpenCL development platform for the entire system.
To test performance the SiSoftware OpenCL GPGPU benchmark suite runs computationally intense algorithms like the Mandelbrot set. Working with SiSoftware, AMD has optimized the performance of the OpenCL benchmarks for its GPU implementations, and for some problems has demonstrated significant performance advantages using AMD’s ATI Stream Software Development Kit (SDK) for OpenCL. When compared to NVIDIA’s CUDA running on its GeForce GTX 295 featuring two GPUs, the ATI Radeon HD 5870 graphics card with one GPU delivers up to 2.7 times faster performance on certain benchmark tests. For the “native float shader” results, the ATI Radeon 5870 posted a score of 1820 megapixels per second, compared to the GTX 295 at 680 megapixels per second. AMD and SiSoftware are currently collaborating on measurement for AMD’s entire platform, including x86 CPUs.
More information over here