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 : Advanced Micro Devices - Moderated (AMD) -- Ignore unavailable to you. Want to Upgrade?

To: Gopher Broke who wrote (20180)	11/22/2000 10:19:46 AM
From: combjelly	Read Replies (1) \| Respond to of 275872

"Weren't there some rumours about a longer pipe?" Yes, although most of those rumors were within the context of Hans DeVries' speculation on AMD patents that would allow the removal of a pipeline stage or two. I suspect that there just wasn't time to make those kind of changes, but that leaves the possibility of a better branch prediction and/or hardware pre-fetch, which are buried quite so deep in the core.

To: Gopher Broke who wrote (20180)	11/22/2000 10:28:39 AM
From: jcholewa	Read Replies (3) \| Respond to of 275872

> What about the opposite? No, not Jim being a monkeys nephew, I mean enhancements to increase > clock speed, possibly at the expense of ipc? Weren't there some rumours about a longer pipe? At this point, I am strongly doubting that there have been enhancements made to the instruction pipe. I believe Jerry is currently happy with the size of his pipe, and doesn't need to artificially enlarge it to please the wome .. er .. the consumers. <g> I would have liked just a teensy change in the direction of pipeline lengthening, but as it is the process boost will probably help enough for the time being for AMD to compete. I should note that it appears as if AMD's ramping problems are power/heat related, not speed path related (as was the problem with the PIII at >1.00GHz), so the particular process changes they are making should be sufficient for the time being. Incidentally.... I personally believe that the Athlon at 1.20GHz is generally similar in speed or faster than the Pentium 4 at 1.50GHz. If you take this as a certainty in this perhaps not so hypothetical scenario, how would you think 1.50GHz and 1.60GHz Athlons would perform against the expected-at-Q3 2.00GHz Pentium 4? I'm asking because there are differences in scaling which should be noted. Would the higher memory bandwidth of the i850 allow the Pentium 4 to scale better as frequency increases, or would the lower start frequencies and greater execution resources allow the Athlon to scale better? Mind you, for this intellectual exercise I am pretending that the Athlon goes in the way Anand suggests, with the Mustang core not having any per-clock performance improvements. Just curious. :)     -JC