Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: H7 using a g3/233 for the match?

Author: Richard A. Fowell (fowell@netcom.com)

Date: 20:58:39 03/20/99

Go up one level in this thread


On March 19, 1999 at 00:25:42, Bruce Moreland wrote:

>
>On March 18, 1999 at 18:51:20, Micheal Cummings wrote:
>
>>
>>On March 18, 1999 at 18:24:22, Will Singleton wrote:
>>
>>>
>>>Matt,
>>>
>>>I see you'll be using a g3/233.  From my experience with H6 on the G3 series, I
>>>find it benefits quite a lot from faster hardware.  You might pick up quite a
>>>rating boost if you were to use a g3/400 (main bus at 100mhz(?) vs 66 for the
>>>older 233).
>>>
>>>Have you considered renting/leasing a g3/400 for the match?  Should'nt be very
>>>expensive.
>>>
>>>Will
>>
>>
>>I agree, I would have thought if you were going to play a game like this and
>>make it a big deal for the chess community that you would try and go for the
>>fastest PC on the market, maybe not the new PIII chips but at least a PII 450.

If Matt can talk Yermolinsky into allowing a faster machine, that would
give HIARCS better chances. But it should still be a G3.

The fastest PC on the market (as far as HIARCS 7.0 is concerned) is the G3.

The stock Apple G3/400 is as fast as any current PC, and several 3rd parties
are selling G3/466 CPU upgrades. Further, the 466 takes well to overclocking
(unlike the PIII), and at least one 3rd party card defaults to 550 MHz when
put in the new Apple G3/300/350/450. (http://www.macintouch.com/g3zif466.html)

Here's my numbers - see my data at the end.

Matt's       G3/233 = Pentium II/320 MHz
Stock, Apple G3/400 = Pentium II/557 MHz
3rd party    G3/466 = Pentium II/650 MHz
Overclocked  G3/550 = Pentium II/750 MHz


 I
>>think Hiarcs 7 will get its butt whipped. I think the hardware used will make
>>this a pretty easy kill for the human player.
>
>How fast is a Power PC 233 thing when compared to a Pentium II?  What is the
>approximately equivalent Pentium II chip?
>
>bruce

Power PC 233 Mz = Pentium II 320 MHz,

Data (note that this data is based on 2.5 hours of testing on each platform,
over 33 problems chosen to include tactical, positional, and endgame
situations, and at time control representative of the upcoming match).

HIARCS 7 - with 64Mb Hash (Mac) or 63Mb Hash (PC)

Machine       CPU MHz CPU   Nodes/second nps/MHz     Configuration
===================================================================
Mac  Yosemite  400 MHz 750   53,413 nps   133.5  Sys 8.5.1, 128Mb RAM, G3
Mac  Yosemite  350 MHz 750   47,615 nps   136.0  Sys 8.5.1, 128Mb RAM, G3
Mac  PowerBook 300 MHz 750   40,062 nps   133.5  Sys 8.5,   192Mb RAM, G3
Mac  ???       310 MHz 750   38,929 nps   125.6  Sys 8.1,   accel. card?
Pent II  /400  400 MHz PII   38,347 nps    95.9  Win 98 "safe mode" (DOS)
AMD  K6/350    350 MHz K6    35,052 nps   100.1  Win 95, -x 63 Mb hash
Mac  iMac      233 MHz 750   30,942 nps   132.8  Sys 8.5.1,  96Mb RAM
Mac  7300/180  180 MHz 604e  20,567 nps   114.3  Sys 7.5.5,  80Mb RAM
Pentium MMX    200 MHz P5?   18,910 nps    94.5  SSDF test machine, 64Mb RAM
Mac  6100/ 66   66 MHz 601    4,862 nps    73.7  Sys 7.5  ,  72Mb RAM





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.