To: Dave B who wrote (25196 ) 7/19/1999 7:10:00 PM From: Tom Warren Read Replies (4) | Respond to of 93625
Intel chose to officially disclose it's PC133 evaluation activities directly to the trade press because they felt speculation would be worse than the reality. They are probably right. After a fairly lengthy discussion with a source close to the matter, I am left with several strong impressions; first that Intel does not have an adequate contingency plan for drdram failing to meet its price and volume goals by mid 2000, second VIA is the cause for re-thinking that position, third in order to develop a contingency Intel would have to interact with enough other companies that they couldn't possibly do it without some public position as to why, fourth that this has nothing to do with the camino rollout or the performance market or the long term desirability of drdram. Looking at each of these in turn. First the lack of contingency. PC100 was the contingency and that was adequate so long as PC133 was not an option. Frankly Intel has behaved all along as though drdram couldn't miss. The recent freefall of sdram may have changed the equation at least temporarily. Drdram is not more costly than thought but because sdram has fallen so sharply the relative price differential may be greater than anticipated. Second, VIA has repositioned itself as more than a chipset designer, by the intended deal with NSM. They may have promoted themselves into the hot seat. Intel would not leave a segment for VIA to operate unopposed even if that segment was unattractive to Intel in a vacuum. Put another way, Intel had no need for a PC133 chipset without VIA manufacturing one. If Intel cannot stop VIA through legal action ( which I expect will be vigorous) they may choose to produce a competitive chipset to deny VIA as much revenue as they can. Intel cannot explore a PC133 alternative in a vacuum. Rather than let speculation rule the news, Intel disclosed the fact they are evaluating PC133 while reaffirming camino, Intel's relationship with Rambus, the advantages of drdram, and its expectation drdram will become dominant in PCs Finally, this isn't about drdram IMHO its about competing with VIA. Intel will compete with rambus vs PC133 but they should have a contingency to compete Intel/PC133 vs VIA/PC133 as well. If they use it, it will be short lived, delaying drdram in the low end by 1 or 2 quarters.