Author: Tom Kerrigan
Date: 15:16:17 09/03/01
Go up one level in this thread
On September 03, 2001 at 12:13:47, Vincent Diepeveen wrote: >On September 02, 2001 at 17:41:04, Tom Kerrigan wrote: > >>On September 01, 2001 at 11:05:29, Gian-Carlo Pascutto wrote: >> >>>On September 01, 2001 at 09:49:44, Charles Young wrote: >>> >>>>There is an independent article the the P4 3.5 ghz exists. Surely, if amd had >>>>such a chip independent writers would have columns on it. >>> >>>Those 'independent writers' get their information directly from Intel. >>> >>>What Vincent means is that while you can demonstrate a cool technology, >>>that doesnt mean you'll be able to deliver it on the desktops for the >>>near future. >> >>In this case, it does. If Intel can make a chip that runs at 3.5GHz, that means >>they just have to get their process running well enough to produce enough for >>commercial purposes. That takes a finite amount of time. When did Intel demo the >>2GHz P4? Less than a year ago, I thought. And now we can buy them. > >Alpha had a 1Ghz chip in lab around 1997. >In 1997 they had a 633Mhz 21164 overclocked to 733Mhz for >2 programs in the wmcc (october 1997). > >The fastest production alpha 21264 now is 833Mhz now? This is a political issue. I believe the 21264/833 is still an EV67, which means that its feature size is 0.25 microns. The P3 and K7 topped out at about 600MHz with similar processes, so it's a testament to the 21264's design that it can hit 833MHz with a shorter pipeline than either. If the project could just get the people and money behind it to put a 0.18 um (or even 0.13 um) part into mass production, it would easily be faster than any PC processor. But now that Intel has essentially cancelled Alpha, I doubt we'll ever see such a thing. >The P4 you can easily overclock of course by clocking back it's >L2 cache (not to mention the trace cache of it), >the cpu then slows down a factor 2 also if not more. Do you have any references to indicate that it's possible to underclock the P4's L2 cache relative to the core? I've read many, many reviews of the P4, involving a lot of overclocking (which is easily done without changing any cache speeds), and never read anything about slowing the L2. >A 1.2Ghz K7 MP outperforms a 2Ghz P4 right now at nearly >every aspect, that says quite something... Bull****. http://www4.tomshardware.com/cpu/01q3/010827/index.html The P4/2GHz beats the K7/1.2GHz on every single benchmark in this review, and sometimes by a huge margin (look at the MPEG4 encoding numbers). >The MP athlon is square, the tbird is rectangular in size, so it's >a completely different design! Do you also think that when a car company changes the wheel size of a particular model, that's a complete redesign of the car? Give me a break. There are dozens of web sites out there that will explain the differences between a Palomino and a Tbird. -Tom
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.