Do Intel's delays hint of 45nm woes?

Posted by Rick C. Hodgin

Indianapolis (IN) - There's an article over at The Inquirer by Charlie Demerjian that I think should be read by many.  In that article, Demerjian states hearing little rumblings from off-the-record sources, those confirming problems with either Intel's 45nm process technology or products.  While it is definitely nothing concrete, it's also not nothing.

Intel recently pushed back the introductory date of three quad-core processors by several weeks (see TGDaily.com's coverage).  While they haven't officially changed any previously announced timeframes, they were able to wriggle into a few more weeks without having to go on the record as to why there was any kind of delay.  In fact, Intel refused to comment when TG Daily explicitly asked them for an explanation for the delay, citing only that there was no delay and that they were still on track for a Q1 08 launch, which was always their announced intention (here).

According to Demerjian, and an objective outsider's view of the minor change in release dates, there's something going on right now.  It's enough to suggest Intel has backpedaled a portion of its 45nm product launch in some ways.  Demerjian also adds that an entire run of Harpertowns had to be scrapped, which were quad-core Xeons.  If true, other questions relating to the scrapping arise.

Demarjian sums it up this way with this closing comment, "There are too many things coming in from too many places for me to think all is well, but nowhere near enough for me to believe that anything is definitively wrong."

Author's opinion
Intel wouldn't move release dates without some valid reason.  Is Intel just taking advantage of AMD's current heartaches--responding, as it were, the way a company should when the unexpected pressure relief valve opens by their competition's folly?  Or, is it something more?