Author: Vincent Diepeveen
Date: 05:08:29 04/26/01
Go up one level in this thread
On April 25, 2001 at 11:33:09, Joshua Lee wrote: >And just think pretty soon the Alphas will be out with 833Mhz and 4Megs of DDR >cache and eventually 1Ghz. I figure that if a 600Mhz DEC is atleast as good if >not better than a 1.5Ghz p4 the 833 and 1Ghz chips will be like a 2 or 3Ghz >Intel or AMD! By then with that speed some chess programs will be able to see >far enough ahead that they may not need a book at best and at the very least be >able to make book moves by calculation. Please calm down a bit. For non-bitboard programs a 21264 at 1Ghz will potentially never get faster as a 1.33Ghz K7 of course. It does 4 instructions a clock, versus K7 3. So even if it has luck with better branch prediction techniques (though those of the K7 are similar to 21264) then the potential luck will not get over 25% soon. Note that the branch misprediction penalty of the 21264 is *way* bigger as that of the P3 or K7. It's more near the same penalty as the P4 has. P4 is an unfinished processor, i would never buy it.
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.