Intel Stacks The Deck?
There's an interesting stink in the PC Testing community. It appears that the SYS2002 benchmark (a yearly benchmark used to compare performance of computers) was designed to favor Intel. In fact it appears that Intel had control over how the benchmark was designed. Van's Hardware (which in the past has been accused of being overly supportive of AMD) has a very well reasoned and presented case on how the benchmark is slanted.
One example is that the Photoshop filters that favor Intel are repeated, thus giving them more weight then the filters AMD is superior in. In other cases, test that favored AMD were removed, while test that favored Intel were increased in weighting. It also appears that weighting is slanted in favor of memory bandwidth, one of the few real advantages that Intel has over AMD. More details can be found in a document [PDF ] released by AMD. What stands out is that AMD took two machines (one AMD, one Intel) and tested them twice (once using SYSMark2001 and once with SYSMark2002). On the 2001 benchmark the AMD system had the advantage, while the 2002 test pushed the Intel system into the lead. The swing was very dramatic and should raise some eyebrows.
It'll be interesitng to see how this plays out now that AMD has joined the board of BAPCo (the creators of SYSMark). It should also put a black eye on Intel for their poor ethics in this case...mj
1:01:15 PM 
|
|
