AMD against SYSmark 2012, is a useless benchmark - Before you buy a new processor, many users read the reviews posted online, they face the various models on the market through the benchmarks , or software that measure the performance when running multiple applications notes. According to AMD , however, SYSmark 2012 (SM2012) is not suitable to evaluate the real performance of the APU and its for that reason, decided not to support more than the benchmark and exit the BAPCo consortium that develops the software.
With a press release published on the company website, AMD has explained the reasons for its decision. Nigel Dessau, senior vice president and head of marketing at AMD, said:
The technology is evolving at an incredible pace, and users need clear and reliable measurements to understand the performance and the value of their systems. AMD believes SM2012 achieves this goal. So AMD can not endorse or support or remain in the consortium BAPCo SM2012.
According to the company, a benchmark should evaluate all components of a system and not to highlight only the CPU performance. SYSmark 2012 uses 18 measurements and 390 applications, but in reality only 7 applications and less than 10% of the measurements determine the total score. In addition, none of the applications included is able to take advantage of GPU acceleration built in APU . The test results are biased and therefore users may be led to make the wrong purchasing choices.
AMD says it is not clear, but it seems that the only benchmark favors Intel CPUs . For the same reason and NVIDIA no longer part of the consortium. The future goal of AMD will now develop a more open and transparent benchmark to give consumers the opportunity to choose a processor based on actual performance.
For Further Reading,
