Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Intel Corporation (INTC) -- Ignore unavailable to you. Want to Upgrade?

To: Dan3 who wrote (139169)	7/13/2001 11:20:05 AM
From: fingolfen	Read Replies (2) \| Respond to of 186894

You are comparing AMD's old core (thunderbird) with Intel's newer core (Willamette). There will never be a 1.6GHZ K7 with the core you are considering. The Palomino core is already in production for mobile and MP versions of Athlon and has design changes that support better scaling such as hardware prefetch and some others (larger BTB size? - I don't recall off hand). They are the current top-end desktop processors sold by Intel and AMD respectively, so it's a fair comparison. Once the Pally makes it to the desktop market, it will supplant the t-bird. Of course, right now it looks like Pally won't hit the desktop market until September, and Willamette will be supplanted by Northwood in October/November. Therefore t-bird/Willamette is a fair comparison, as is Pally/N-wood. But the bottom line is that you need to go back look at the IPC scaling of Palomino if you want to extrapolate AMD's 1.6 or 1.7GHZ performance - because that's the core that will be used in those chips. ...and by the same token I would have to look at the scaling of the Northwood... Cadalyst Magazine just released a bunch of tests comparing a 1.2GHZ Athlon (palomino core) with a 1.7GHZ Xeon and found them to be basically tied in performance. cadalyst.com So a 1.6GHZ Athlon would be expected to perform about the same as a 2.5GHZ P4 - depending on scaling, but preretch etc. is designed to let the newer core scale better than the old one that you used in your analysis. Sorry, don't buy it. The analysis will have to be re-run once we have data for Pally and N-wood over multiple speed grades...