Author: Robert Hyatt
Date: 12:08:13 05/24/04
Go up one level in this thread
On May 24, 2004 at 14:12:10, Slater Wold wrote:
>On May 24, 2004 at 14:09:56, Robert Hyatt wrote:
>
>>On May 24, 2004 at 12:56:56, Slater Wold wrote:
>>
>>>On May 24, 2004 at 12:29:04, Russell Reagan wrote:
>>>
>>>>On May 24, 2004 at 12:06:46, Slater Wold wrote:
>>>>
>>>>>In a thread below, there is a post about Blargh, ICCs top computer. It has won
>>>>>60% of its games at standard time controls, and only lost 16% in almost 3,000
>>>>>games!
>>>>>
>>>>>I am interested to know the OoO (opinion of outcome) of EVERYONE at the
>>>>>following experiment, I will be in contact with the operator of Blargh, and will
>>>>>try to set this up.
>>>>>
>>>>>Crafty 19.13 on an AMD FX53 (64-bit) vs Shredder 8 on a Dual Opteron 2.2Ghz
>>>>>(32-bit)
>>>>>
>>>>>10 game match. 120/0.
>>>>>
>>>>>Opinions?
>>>>
>>>>Crafty on equal hardware probably doesn't stand a realistic chance against
>>>>Shredder. Only Fritz scores at least 45% against Shredder on the latest SSDF.
>>>>Junior and Hiarcs score at least 40%. Crafty would be behind those engines on
>>>>equal hardware by a non-trivial margin (take the Junior-Crafty 10x experiment
>>>>for example).
>>>
>>>OoO please...
>>>
>>>>What is interesting about a Shredder-Crafty match where Crafty is at a hardware
>>>>_disadvantage_?
>>>>
>>>>BTW, that really explains a lot about why he is #1 on ICC. You have a freakin'
>>>>FX53 and you're at a hardware disadvantage! Wow...
>>>
>>>It's a disadvantage, but a small one. How small?
>>
>>It is at least 2x disadvantage.
>>
>>You did notice that box is a _dual_???
>
>Dual. And 32-bit.
>
>>Shredder has an SMP version which I would assume he is using...
>>
>>>Crafty on a 2.2Ghz Opteron in 32-bit gets about 1.5M NPS. With a 1.89 speedup,
>>>we get 2.8M nps. My FX53 (64-bit) will do 2.6M. It's at a 8% HW disadvantage.
>>>
>>>8% means nothing in a 120/0 game.
>>
>>Where are those numbers coming from?
>>
>>
>>Crafty on a 2.2ghz opteron gets >2M on linux with GCC. With windows it is
>>closer to 2.3M on a single opteron CPU...
>
>My 2.4Ghz FX doesn't get 2.3M 32-bit. Where are you getting these numbers?
Running in the last CCT the bench command was doing 2M. I posted the numbers
here at some point. My log files for the DTS positions (on my ftp machine in
the smpdata directory) shows that the first DTS position starts at 1.9M nodes
per second. Here are the actual one-cpu runs:
medusa% grep nps cbrun*0
time=5:00 cpu=99% mat=-3 n=570081156 fh=92% nps=1.90M
time=5:00 cpu=99% mat=-3 n=572380305 fh=93% nps=1.91M
time=5:00 cpu=99% mat=-2 n=601305165 fh=92% nps=2.00M
time=5:00 cpu=99% mat=-2 n=665116633 fh=95% nps=2.22M
time=5:00 cpu=99% mat=-1 n=650939746 fh=94% nps=2.17M
time=5:00 cpu=99% mat=-2 n=657190763 fh=93% nps=2.19M
time=5:00 cpu=99% mat=-2 n=671293016 fh=93% nps=2.24M
time=5:00 cpu=99% mat=1 n=649961387 fh=92% nps=2.17M
time=5:00 cpu=99% mat=-1 n=641706498 fh=91% nps=2.14M
time=5:00 cpu=99% mat=-1 n=635992282 fh=92% nps=2.12M
time=5:00 cpu=99% mat=0 n=687746721 fh=92% nps=2.29M
time=5:00 cpu=99% mat=0 n=697967569 fh=92% nps=2.33M
time=5:00 cpu=99% mat=0 n=682278911 fh=92% nps=2.27M
time=5:00 cpu=99% mat=0 n=680641307 fh=91% nps=2.27M
time=5:00 cpu=99% mat=0 n=705501897 fh=92% nps=2.35M
time=5:00 cpu=99% mat=5 n=711992860 fh=91% nps=2.37M
time=5:00 cpu=99% mat=0 n=680627671 fh=91% nps=2.27M
time=5:00 cpu=99% mat=0 n=684408845 fh=91% nps=2.28M
time=5:00 cpu=99% mat=-1 n=672072349 fh=91% nps=2.24M
time=5:00 cpu=99% mat=-1 n=639020200 fh=89% nps=2.13M
time=5:00 cpu=99% mat=-1 n=626508871 fh=88% nps=2.09M
time=5:00 cpu=99% mat=-1 n=652655813 fh=89% nps=2.18M
time=5:00 cpu=99% mat=-1 n=646700868 fh=89% nps=2.16M
time=5:00 cpu=99% mat=-1 n=644620982 fh=88% nps=2.15M
As you can see, I am _not_ guessing here...
Here are dual numbers:
time=5:00 cpu=199% mat=-3 n=1125716661 fh=92% nps=3.75M
time=5:00 cpu=199% mat=-3 n=1143745837 fh=93% nps=3.81M
time=5:00 cpu=199% mat=-2 n=1170889476 fh=93% nps=3.90M
time=5:00 cpu=199% mat=-2 n=1295137485 fh=95% nps=4.32M
time=5:00 cpu=199% mat=-1 n=1306020483 fh=94% nps=4.35M
time=5:00 cpu=199% mat=-2 n=1312034042 fh=93% nps=4.37M
time=5:00 cpu=199% mat=-2 n=1334565314 fh=94% nps=4.45M
time=5:00 cpu=199% mat=1 n=1275333138 fh=92% nps=4.25M
time=5:00 cpu=199% mat=-1 n=1252428376 fh=91% nps=4.17M
time=5:00 cpu=199% mat=-1 n=1263898857 fh=92% nps=4.21M
time=5:00 cpu=199% mat=0 n=1344520663 fh=92% nps=4.48M
time=5:00 cpu=199% mat=0 n=1360924345 fh=92% nps=4.54M
time=5:00 cpu=199% mat=0 n=1322792409 fh=91% nps=4.41M
time=5:00 cpu=199% mat=0 n=1313789012 fh=91% nps=4.38M
time=5:00 cpu=199% mat=0 n=1339933480 fh=91% nps=4.47M
time=5:00 cpu=199% mat=5 n=1341070056 fh=91% nps=4.47M
time=5:00 cpu=199% mat=0 n=1302945219 fh=91% nps=4.34M
time=5:00 cpu=199% mat=0 n=1309092015 fh=91% nps=4.36M
time=5:00 cpu=199% mat=-1 n=1281171538 fh=90% nps=4.27M
time=5:00 cpu=199% mat=-1 n=1234130301 fh=89% nps=4.11M
time=5:00 cpu=199% mat=-1 n=1203620866 fh=88% nps=4.01M
time=5:00 cpu=199% mat=-1 n=1257942664 fh=88% nps=4.19M
time=5:00 cpu=199% mat=-1 n=1257366699 fh=88% nps=4.19M
time=5:00 cpu=199% mat=-1 n=1253818748 fh=88% nps=4.18M
And here are some 4-cpu numbers:
time=5:00 cpu=397% mat=-3 n=2214294718 fh=92% nps=7.38M
time=5:00 cpu=398% mat=-3 n=2275918132 fh=93% nps=7.59M
time=5:00 cpu=398% mat=-2 n=2324669850 fh=93% nps=7.75M
time=5:00 cpu=354% mat=-2 n=2569040102 fh=95% nps=8.56M
time=5:00 cpu=398% mat=-1 n=2483899624 fh=93% nps=8.28M
time=5:00 cpu=396% mat=-2 n=2524386780 fh=93% nps=8.41M
time=5:00 cpu=398% mat=-2 n=2625619952 fh=93% nps=8.75M
time=5:00 cpu=399% mat=1 n=2533565587 fh=92% nps=8.45M
time=5:00 cpu=399% mat=-1 n=2471614867 fh=91% nps=8.24M
time=5:00 cpu=398% mat=-1 n=2485704184 fh=92% nps=8.29M
time=5:00 cpu=398% mat=0 n=2669553594 fh=91% nps=8.90M
time=5:00 cpu=398% mat=0 n=2695287412 fh=92% nps=8.98M
time=5:00 cpu=397% mat=0 n=2552217576 fh=91% nps=8.51M
time=5:00 cpu=398% mat=0 n=2584530982 fh=91% nps=8.62M
time=5:00 cpu=398% mat=0 n=2641396390 fh=91% nps=8.80M
time=5:00 cpu=397% mat=5 n=2620717407 fh=91% nps=8.74M
time=5:00 cpu=397% mat=0 n=2568920255 fh=91% nps=8.56M
time=5:00 cpu=396% mat=0 n=2542256969 fh=91% nps=8.47M
time=5:00 cpu=396% mat=-1 n=2494151402 fh=90% nps=8.31M
time=5:00 cpu=396% mat=-1 n=2388262254 fh=88% nps=7.96M
time=5:00 cpu=394% mat=-1 n=2333202169 fh=88% nps=7.78M
time=5:00 cpu=395% mat=-1 n=2451027666 fh=88% nps=8.17M
time=5:00 cpu=396% mat=-1 n=2475641449 fh=88% nps=8.25M
time=5:00 cpu=396% mat=-1 n=2446281739 fh=88% nps=8.15M
No guesswork required...
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.