Hardly surprising...
Lately there\'s an article at o/cers.com saying Intel is always relunctant (and sometimes too late) to correct problems as a means to cut cost...
http://www.overclockers.com/tips01336/引用:
Yes, changes in gross margin percentage at a company like Intel do affect real profits more than in most companies, so there\'s some legitimate reason for the obsession.
But that obsession ends up putting tremendous pressure on Intel to avoid "extra" expenses to fix problems that could hurt that beloved measurement short-term. The problem is the problems don\'t go away, eventually they become unavoidable, and Intel ends up on the sidelines for a while, giving AMD another chance to shine for a while.
That\'s essentially why Intel ran the PIII and PIV designs into the ground. They knew they had heat problems with both, and they knew really fixing the problem would be very expensive. So they didn\'t. When the PIII blew up, they ducked the problem with the PIV design, and they didn\'t bite the bullet until Prescott nearly melted on them.
IMO it\'s quite conceivable... Just think how many snafus happened to them in a decade which could be have been avoided...
They knew about the pipeline stall problem in P6-core for years and when did they finally got it fixed? (not before coppermine), Coppermine cB0 1.13GHz recall incident (fixed by release of new stepping), and Pentium 4 (which they could have scrapped the whole line at the release of P-M)...
