Replies to post #39050 on Intel Corp (INTC)

[Elmer Phud]

Smooth -

I've listened to the call and I must say that I am struck by his subdued, almost sheepish tone. I don't think I've ever heard a CEO with such a lack of confidence and leadership in his manner. He really doesn't want to be there and he REALLY doesn't want to answer these questions.

When someone says "I'm really thrilled" in a soft tone and is practically looking down and shuffling his feet you know something is really wrong here...

[alan81]

I am glad he put it in context.
In the same paragraph, he said the current opteron wins as many benchmarks as it loses to core 2. If we use that as the standard for what he is saying about Barcelona, then, well...
--Alan

[VBG]

Need to keep things in context.

Remember, he said "in terms of performance metrics that are valuable,that are important, that are significant and relevant". Valuable, important, significant and relevant to whom?

It may be that "all of those benchmarks" = 2
( SpecFP2006_rate and Stream )

I'm curious about what they "anticipate seeing in the market".
IIRC, Intel has publicly stated 3GHz Clovertown w/ 1.6GHz FSB and a Penryn based Xeon of undisclosed speed both in H2'07

-VBG

[wbmw]

Re: "The next generation of product that we are introducing here, the real quad core, this is really a significant jump in all of those benchmarks, and we are pretty excited and pretty thrilled that we are looking at numbers of 3 and 4 decades of points, you know 30-40% improvement from what we have seen in the market and anticipate seeing in the market."

Interesting. Just last week he was claiming 42% performance advantage, and now it's a 30-40% advantage. With Clovertown 3.0GHz, it will be a 20-30% advantage... in the SPECfp_rate and the other "important" AMD benchmarks. Overall, it will probably be a case of -20% to +30% performance variation. At least until 1600 MT/s FSB, 45nm, and other Aces that Intel has up their sleeves.