To: niceguy767 who wrote (37414 ) 4/26/2001 1:27:01 PM From: dale_laroy Read Replies (1) | Respond to of 275872 "It's difficult, it seems, to accept the new paradigm whereby between now and 2 years from now, it is highly likely that AMD will advance at a quicker pace than will INTC no matter what delays both suffer during the interim...i.e 2 years out AMD will have widened the existing gap!!!" If you had not included that, "no matter what delays both suffer" I would definitely agree. Indeed, I would agree if you really meant both instead of each. With both vendors having the same delays, such as migration to 0.13-micron equipment, AMD is definitely favored, for the same reason that many of AMD's potential delays can not be suffered by Intel. Intel can not have either their transition to x86-64 or SOI delayed until 2003, because neither is on Intel's roadmap at all. 1) Palomino is a major introduction. It is more significant than either Tualatin, and even Northwood, at least until Intel can get Northwood volume up. Palomino gets AMD into two markets, mobile and workstation, that are currently essentially Intel only. 2) While Intel will initially get more speed grade improvement frpm the migration to 0.13-micron, due to their migration from aluminum to copper, AMD will gain more with the migration to SOI. 3) The migration to 0.13-micron will increase AMD's unit volume by over 50%. AMD can use this to increase market share. Intel will find an increase in market share extremely difficult. 4) Intel has nothing to compare to Hammer on the horizon. Offsetting this is several potential negatives: 1) If the move to SOI is delayed until 2003, Intel will pull even further ahead of AMD's clock rate, especially since Hammer is exclusively SOI. 2) Any delay of Hammer to 2003 could be very costly. While SOI 0.13-micron Thoroughbred will compete better against Tualatin/Northwood than Palomino versus Tualatin/Willamette, a delay of Hammer into 2003 would almost assure AMD of minimal penetration into the enterprise systems market. 3) Unless Thoroughbred has SSE(2), the shift of the software base in H2 2002 would make a delay of Hammer till 2003 very costly.