Author: Ricardo Gibert
Date: 03:28:29 01/10/02
Go up one level in this thread
On January 10, 2002 at 06:09:04, Gian-Carlo Pascutto wrote: >On January 10, 2002 at 06:05:42, Ricardo Gibert wrote: > >>On January 10, 2002 at 05:45:23, Gian-Carlo Pascutto wrote: >> >>>On January 10, 2002 at 05:37:00, Ricardo Gibert wrote: >>> >>>>On January 10, 2002 at 03:07:44, Gian-Carlo Pascutto wrote: >>>> >>>>>On January 10, 2002 at 02:59:03, Ricardo Gibert wrote: >>>>> >>>>>>On January 07, 2002 at 16:00:58, Gian-Carlo Pascutto wrote: >>>>>> >>>>>>>On January 07, 2002 at 12:53:17, Rajen Gupta wrote: >>>>>>> >>>>>>>>i'm afraid that amd may lose all the goodwill they have earned so far unless >>>>>>>>they get their act together real soon >>>>>>> >>>>>>>Funny, I've been (and still am) thinking the same of Intel >>>>>>>for quite a while. >>>>>>> >>>>>>>The P4 is garbage from a technical point of view, and >>>>>>>their current Itaniums seem to be too, based on the benchmarks >>>>>>>we have seen so far. Their only hope is to keep pushing up >>>>>>>the clockspeed as high as they can, but the design has limits >>>>>>>*somewhere*, as the original P3 at 1.13Ghz showed (totall recall >>>>>>>due to instabily). >>>>>> >>>>>>http://www.theregister.co.uk/content/3/23593.html >>>>> >>>>>Please read what I wrote. I said the _original_ P3. >>>> >>>>Please read what you wrote. You said, "..., but the design has limits >>>>*somewhere*, as the original P3 at 1.13Ghz showed..." >>> >>>Right. So with smaller process technology, it can go a bit higher, >>>but it's still limited. >> >>Limited? I require proof to accept such a statement. Have you seen such proof or >>are you guessing? >> >>I know what you are saying has been alleged in many places on the internet, but >>I have yet to read an explanation (much less see a proof) of how that can be so. >>It's counter-intuitive and therefore requires compelling evidence. > >A good indication (not proof) would be that so far all CPU's Intel produced have >reached a platform above which they got unstable. That platform can be upped by >smaller process technology. But you cannot keep decreasing process technology, >as there is a lower limit there. (single atoms sounds like a good lower bound) > >-- >GCP Ah, yes, there *is* that limit, but all designs have that limit. Before that limit, heat is another likely limitation regardless of die shrink. Some physicist came up with an explanation of how this might be so. But the issue is whether the P3 had hit some kind of limit and more likely the limit was an artificial one. Intel simply wanted to put their money on the P4 rather than the P3 as far as desktops are concerned. So far, I don't see how they were necessarily wrong to do so.
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.