Author: Slater Wold
Date: 06:08:22 05/30/02
Go up one level in this thread
On May 30, 2002 at 08:33:52, Vincent Diepeveen wrote: >On May 30, 2002 at 01:23:35, Slater Wold wrote: > >>On May 29, 2002 at 20:34:37, Scott Gasch wrote: >> >>>On May 29, 2002 at 19:50:20, Ricardo Gibert wrote: >>> >>>>On May 29, 2002 at 18:39:03, Scott Gasch wrote: >>>> >>>>>I've been gently informed that I'm not supposed to talk about the speed of >>>>>not-released processors, sorry about that. >>>>> >>>>>Scott >>>> >>>>I have to admit I realized you were unwittingly violating an NDA, but my >>>>curiosity got the better of me. I should have told you straight off what you >>>>were disclosing might get you in trouble. My apologies. >>> >>>No apologies needed, it's my own stupidity at fault here. For the record I'm >>>not in any trouble. I self-censored when it was pointed out to me that I could >>>possibly get into trouble with this. >>> >>>Scott >> >>Well hell, now you got us all curious. Now I wanna know what was said. ;) > >Well i got info from deep throat lucky. Other source in this case. >This source confirms that for 64 bits applications like crafty the >i2 is very fast. 1.5 million nodes a second is not bad. > >your bench you posted here at 2.53Ghz P4 was like 950k a second or so? > >that's of course major victory for intel getting 1.5 MLN nodes a second >on a 1Ghz processor. P4 2.53Ghz was 965k nps. Profiled P4 2.53Ghz was 1.09M nps. Profiled AMD 1.73Ghz was 1.0M nps. Profiled dual AMD 1.73Ghz was 1.6M nps. So one i2 is == to a Dual AMD 1.73Ghz. (IN TERMS OF SPEED.) And those suckers are *very* scalable. (While AMDs are not ATM.)
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.