Q&A On An AMD Delay . . . 1794

Somebody asked me about the purported AMD delay in Phenom chips, and this is the short-and-sweet answer:

The initial roadmap talked about “order availability” by X date. These new dates are a quarter later, but it’s not reported if these are order date or actually-around dates. If it’s the second, it’s probably not a “real” delay at all.

My real suspicion, though, is that this is this is just a placeholder until the next last-minute delay. I don’t doubt they’ll manage to get the lower speeds out more or less on time, but the rest?

The real underlying issue, though, is whether they’re inherent as good or better than Penryns clock-for-clock. The leaked Cinebench results indicate that they’re not. If they aren’t, then clock speed is almost irrelevant, and all hell is going to break loose.

P.S. Besides the date, the only real difference between the two was that some lower speeds were added to the list. This may be perfectly normal and unremarkable, or it may be a safety net to be used if the processors can’t be
speed-ramped enough on time. It looks better to get something slow-but-scheduled a while back on time rather than having to invent slow ones on the spot.

It should also be noted that AMD’s status rides solely on its latest stepping and spin, and all else that goes before that becomes irrelevant. It’s the last stepping before going into production that counts. Not like we’ll see a 4GHz chip come out of the next try, but the current picture can brighten considerably almost overnight.

Ed


Be the first to comment

Leave a Reply