AMD, SiSoftware Team to Develop Benchmark Testing Suites for OpenCL

Advanced Micro Devices (News - Alert), or AMD, reportedly announced its involvement in the development of one of the first industry benchmark testing suites for OpenCL.

 

OpenCL is an industry standard for running parallel tasks on CPUs and GPUs using the same code. AMD (News - Alert) provides a complete OpenCL development platform for the entire system.

                         

The developer community, independent software vendors (ISVs) and original equipment manufacturers have been looking for a way to measure OpenCL-based system performance.

 

Company officials said that the OpenCL benchmark suite from SiSoftware is a first step in giving the industry the tools it needs to accurately measure and assess system performance. OpenCL GPGPU benchmark suite is part of SiSoftware Sandra 2010, which is the latest version of the utility that was originally introduced in 1997.

 

In addition to remote analysis, the benchmark suite includes benchmarking and diagnostic features for PCs, servers, mobile devices and networks. Moreover, it can be used to test OpenCL performance on ATI Stream technology.

 

With SiSoftware, AMD said that it has optimized the performance of the OpenCL benchmarks for its GPU implementations. In addition, for some problems, the company has demonstrated performance advantages using AMD ATI Stream Software Development Kit for OpenCL.

 

To test performance, the SiSoftware OpenCL GPGPU benchmark suite runs computationally intense algorithms like the Mandelbrot set. Unlike NVIDIA’s (News - Alert) 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, said company officials.

 

Company officials said that for the “native float shader” results, the ATI Radeon 5870 posted a score of 1820 megapixels per second, compared to that GTX 295 posted a score of 680 megapixels per second.

 

Currently, SiSoftware and AMD are collaborating on measurement for AMD’s entire platform, including x86 CPUs.

                         

Last month, AMD released ATI Radeon HD 5970, a dual-chip high-end card featuring two clock-speed-reduced Radeon HD 5870 GPUs on a single graphics card.