A number of you have been writing in with word that Van?s Hardware has “discovered” that SysMark 2002is rigged in Intel?s favor. BAPCo., the publisher of SysMark has been accused by AMD of not only using Intel software engineers to develop the benchmark, but also of skewing their scoring methods in favor of memory bandwidth intensive results, a key area of Intel prowess (owing to RDRAM, of course). The accusation is based in part on the obvious scoring differences between the pre-Athlon XP SysMark 2001 and the post-XP SysMark2002. Needless to say, these are hefty charges, but about half of you who wrote in dismissed the accusations as mere perpetrations of an AMD fanboy (Van is known for his AMD leanings). Part of the resentment comes from the fact that Van is basically reporting the contents of an AMD PDF (that?she?s hosting, too). The PDF, which has been distributed to a number of sites, constitutes more or less all of his complaint. So, on the one hand, he’s not making this stuff up–AMD is really out pushing this topic. On the other hand, why is AMD pushing this issue in such a seemingly irresponsible way? In any case, more data is needed to determine if in fact the changes to SysMark2002 were puppeted from the towers of Intel, and I suspect that we?ll hear from BAPCo before too long. AMD has said that they would now be joining BAPCo to ensure that this situation does not repeat itself when SysMark2003 comes available.
