Not logged inRybka Chess Community Forum
Up Topic The Rybka Lounge / Computer Chess / FinalGen
1 2 3 4 5 Previous Next  
- - By Dhanish (***) Date 2012-03-19 14:30
Just used FinalGen (http://www.mtu-media.com/finalgen/home_ing.php) to analyze the famous position Pachman-Hromadka 1944, see http://rybkaforum.net/cgi-bin/rybkaforum/topic_show.pl?pid=138330;hl=Pachman%20Hromadka. It took only 36 min on my Pentium IV laptop to give the solution! It appears to be quite fast. Since one can pause the analysis and continue, time does not seem to be a limitation. Anyway, looking forward to further developments.
Parent - By DGB (**) Date 2012-03-19 14:35
Cool.
Parent - - By keoki010 (Silver) Date 2012-03-19 16:14
Did you get any malware errors? I get malware threats when I try to download it. I am using Avast lastest update.
Parent - - By Dhanish (***) Date 2012-03-19 16:29

> Did you get any malware errors?


No, I am using Microsoft Security Essentials.
Parent - - By keoki010 (Silver) Date 2012-03-19 16:36
Ok, ever since I upgraded I've been getting a lot of threats.  They must have gone overboard on their threat detection.  Thanks.
Parent - - By mexicanstandoff (*) Date 2012-04-08 12:12
Overboard on threats??

The DLLs generated by the setup package are infected, period!

Could you please return to the forum and address the issues? I have used your website's contact, too.
Parent - - By Pia (****) Date 2012-04-08 12:21
Where in the hell any of these small DLLs are infected?
Could you show me injected virus/trojan code,
not just antivirus brand names?
As you are programmer, top one.
Parent - - By mexicanstandoff (*) Date 2012-04-09 04:37

>Could you show me injected virus/trojan code, not just antivirus brand names?


Tell you what, I'll snailmail suspected malware authors requesting a certificate of authenticity ("I hereby certify the program with an SHA256 of xyz is a genuine trojan, and here is a digitally-signed eTrust license").  Keep waiting, I'll get back to you with their response.
Parent - By Pia (****) Date 2012-04-09 09:08
Good luck with that. Meanwhile, i've already reported to Comodo for these DLLs as a false positives.
It worked in the past, so you better crack into smaller bits your soft theory.
Parent - By RFK (Gold) Date 2012-04-09 14:12
I have no need to experiment with this -It is too much of a project to go through the process of formatting my hard drive. I'll satisfy myself with Aquarium IDeA.
Parent - - By RFK (Gold) Date 2012-03-20 13:57
Who are the developers?
Parent - By Fulcrum2000 (****) Date 2012-03-20 15:17
From a forum he posted this program on:

I am an endgame enthusiast. When I first knew of Nalimov tablebases I was so impressed that a tool could play chess perfectly, that I decided to develop my own tablebase generator.
Depending on the position caracteristics, it can resolve positions with up to 7 or 8 pieces. Even more for pawn endgames.
In my real job, I work as a computer engineer for Airbus, in Toulouse, France. It took me 4 years to develop this tool in my leisure time.
If you are interested, I invite you to visit my website

http://www.mtu-media.com/finalgen
Try it, it's easy to use, and free!
Pedro PĂ©rez Romero
Parent - By mexicanstandoff (*) Date 2012-04-08 12:12

>Who are the developers?


Very good question. IMO, yet to be answered.
Parent - By Gambit-man (**) Date 2012-03-24 18:28 Edited 2012-03-24 18:33
Seems like a fantastic piece of kit, just what i've been looking for, and have been for years.
I expect it would become even better with the advances in computer technology ie, storage space and processor power etc.
Just as a comparison, my AMD dual-core 2x2.7GHz with 2 Gb ram, my machine took 37 minutes and 59 seconds to solve the Timman-Karpov game.

If i have one small complaint, it's that it's not able to deal with endgames where either side has more than one piece.

Perhaps in the future people could upload their own workings out onto some tablebase site where eventually we get close to full endgame tablebases
Parent - - By hsnksc77 (**) Date 2012-03-27 18:09
New game - Stockfish 120305M
8/4R3/p5kp/2r5/8/3K4/P7/8 b - - 0 1


Analysis by Stockfish 120305M:

1...Ra5 2.Re6+ Kg5 3.Re2 h5 4.Rg2+ Kf5 5.Rf2+ Ke6 6.Re2+ Kd6
  -/+  (-1.33)   Depth: 6/11   00:00:00  7kN
1...Ra5 2.Re6+ Kg5 3.Re2 h5 4.Ke4 h4 5.Rh2 Ra4+ 6.Ke5 Ra3 7.Ke4 h3
  -+  (-1.57)   Depth: 7/15   00:00:00  15kN
1...Ra5 2.Re2 h5 3.Rg2+ Kf5 4.Rh2 Ke6 5.Ke3 Rg5 6.Kf4 Rg4+ 7.Kf3 Rg5
  -/+  (-1.37)   Depth: 8/18   00:00:00  51kN
1...Ra5 2.Re2 h5 3.Rg2+ Kf5 4.Rh2 Kg5 5.Ke3 h4 6.Ke4 Ra4+ 7.Ke5 Ra3 8.Rg2+ Rg3 9.Rh2 h3
  -+  (-1.69)   Depth: 9/18   00:00:00  67kN
1...Ra5 2.Re2 h5 3.Rg2+ Kf5 4.Rh2 Kg5 5.Ke3 h4 6.Ke4 Ra4+ 7.Ke5 Ra3 8.Rg2+ Rg3 9.Rh2 h3
  -+  (-1.69)   Depth: 10/18   00:00:00  70kN
1...Ra5 2.Re2 h5 3.Rg2+ Kf5 4.Rh2 Kg5 5.Ke3 h4 6.Ke4 Ra4+ 7.Ke5 Ra3 8.Rg2+ Rg3 9.Rh2 h3 10.Ke4 Rg4+ 11.Kf3
  -+  (-1.93)   Depth: 11/20   00:00:00  124kN
1...Ra5 2.Re2 h5 3.Rg2+ Kf5 4.Rf2+ Kg5 5.Ke4 h4 6.Rg2+ Kf6 7.Rc2 Ra4+ 8.Kf3 Ke6 9.Rc6+ Ke5 10.Rc5+ Kd6 11.Rh5 Ra3+ 12.Ke4 h3 13.Rh6+ Kc5
  -+  (-1.73)   Depth: 12/25   00:00:00  175kN
1...Ra5 2.Re2 h5 3.Ke4 Ra4+ 4.Kf3 h4 5.Rc2 Kf6 6.Rc6+ Kf5 7.Rc5+ Ke6 8.Rc6+ Kd7 9.Rc2 Kd6 10.Re2 Kd5 11.Rc2 h3 12.Rh2 Rh4 13.a3
  -+  (-1.89)   Depth: 13/31   00:00:00  341kN
1...Ra5 2.Re2 h5 3.Ke4 Ra4+ 4.Kf3 h4 5.Rc2 Kf6 6.Rc6+ Kf5 7.Rc5+ Ke6 8.Rc6+ Kd7 9.Rc2 Kd6 10.Re2 Kd5 11.Rc2 Kd6
  -+  (-1.89)   Depth: 14/31   00:00:00  393kN
1...Ra5 2.Re2 h5 3.Ke4 Ra4+ 4.Kf3 h4 5.Rc2 Kf6 6.Rc6+ Kf5 7.Rc5+ Ke6 8.Rc6+ Kd7 9.Rc2 Kd6 10.Re2 Kd5 11.Rc2 Kd6
  -+  (-1.89)   Depth: 15/31   00:00:00  459kN
1...Ra5 2.Re2 h5 3.Ke4 Ra4+ 4.Kf3 h4 5.Rc2 Kf6 6.Rc6+ Kf5 7.Rc5+ Ke6 8.Rc6+ Kd7 9.Rc2 Kd6 10.Re2 Kd5 11.Rc2 Kd6
  -+  (-1.89)   Depth: 16/31   00:00:00  553kN
1...Ra5 2.Re2 h5 3.Ke4 Ra4+ 4.Kf3 h4 5.Rc2 Kf6 6.Rc6+ Kf5 7.Rc5+ Ke6 8.Rc6+ Kd7 9.Rc2 Ke7 10.Rc7+ Kd6 11.Rc2 Kd5 12.Rd2+ Ke5 13.Rc2 Kd5
  -+  (-1.85)   Depth: 17/31   00:00:00  1465kN
1...Ra5 2.Re2 h5 3.Ke4 Ra4+ 4.Kf3 h4 5.Rc2 Kf6 6.Rc6+ Kf5 7.Rc5+ Ke6 8.Rc6+ Kd7 9.Rc2 Ke7 10.Rc7+ Kd6 11.Rc2 Kd5 12.Rd2+ Ke5 13.Rc2 Rf4+ 14.Kg2 Kd5 15.Kh3 Ra4 16.Rg2
  -+  (-1.85)   Depth: 18/31   00:00:00  1583kN
1...Ra5 2.Re2 h5 3.Ke4 h4 4.Rc2 Kh5 5.Kf4 Ra4+ 6.Kf3 Ra3+ 7.Kf4 Ra5 8.Rd2 Ra4+ 9.Kf3 Kg5 10.Rd5+ Kf6 11.Rd2 Kf5 12.Rd5+ Ke6 13.Rd2 Rc4 14.Re2+ Kf5 15.Kg2 Kg4
  -+  (-1.93)   Depth: 19/33   00:00:00  2746kN
1...Ra5 2.Re2 h5 3.Ke4 h4 4.Rc2 Kh5 5.Kf4 Ra4+ 6.Kf3 Ra3+ 7.Kf4 Ra5 8.Rd2 Ra4+ 9.Kf3 Kg5 10.Rd5+ Kf6 11.Rd2 Kf5 12.Rd5+ Ke6 13.Rd2 Rc4 14.Kg2 Rc3 15.Rd4 Rg3+ 16.Kf2
  -+  (-2.02)   Depth: 20/37   00:00:00  3675kN
1...Ra5 2.Re2 h5 3.Ke4 h4 4.Rc2 Kg5 5.Kf3 Ra3+ 6.Kg2 a5 7.Rc5+ Kf4 8.Rc4+ Ke3 9.Rc2 Kd3 10.Rf2 Ke4 11.Re2+ Kd5 12.Rd2+ Ke5 13.Re2+ Kf4 14.Rf2+ Kg4 15.Rc2 Rg3+ 16.Kh2 Rh3+ 17.Kg2 Ra3 18.Rc4+ Kh5 19.Rc5+ Kg6 20.Rc2 Kf5 21.Rf2+ Kg4
  -+  (-2.06)   Depth: 21/37   00:00:00  5257kN
1...Ra5 2.Re2 h5 3.Ke4 h4 4.Rc2 Kg5 5.Kf3 Ra3+ 6.Kg2 a5 7.Rc5+ Kf4 8.Rc4+ Ke3 9.Rc2 Kd3 10.Rf2 Ke4 11.Re2+ Kd5 12.Rd2+ Ke5 13.Re2+ Kf4 14.Rf2+ Kg4 15.Rc2 Rg3+ 16.Kh2 Rh3+ 17.Kg2 Ra3 18.Rc4+ Kh5 19.Rc5+ Kg6 20.Rc2 Kf5 21.Rf2+ Kg4
  -+  (-2.06)   Depth: 22/37   00:00:00  5954kN
1...Ra5 2.Re2 h5 3.Ke4 h4 4.Rc2 Kg5 5.Kf3 Ra3+ 6.Kg2 a5 7.Rc5+ Kf4 8.Rc4+ Ke3 9.Rc2 Kd3 10.Rf2 Ke4 11.Re2+ Kd5 12.Rd2+ Kc4 13.Rf2 Rg3+ 14.Kh2 Kd3 15.Rf4
  -+  (-2.06)   Depth: 23/37   00:00:01  7050kN
1...Ra5 2.Re2 h5 3.Ke4 h4 4.Rc2 Kg5 5.Kf3 Ra3+ 6.Kg2 a5 7.Rc5+ Kf4 8.Rc4+ Ke3 9.Rc2 Kd3 10.Rf2 Ke4 11.Re2+ Kd5 12.Rd2+ Kc4 13.Rf2 Rg3+ 14.Kh2 Kd3 15.Rf4
  -+  (-2.06)   Depth: 24/37   00:00:01  8354kN
1...Ra5 2.Re2 h5 3.Ke4 h4 4.Rc2 Kg5 5.Kf3 Ra3+ 6.Kg2 a5 7.Rc5+ Kf4 8.Rc4+ Ke3 9.Rc2 Kd3 10.Rf2 Ke4 11.Re2+ Kd5 12.Rd2+ Kc4 13.Rc2+ Kd4 14.Rd2+ Ke5 15.Re2+ Kf5 16.Rf2+ Kg4 17.Rc2 Rg3+ 18.Kh2 Rh3+ 19.Kg2 Ra3 20.Rc4+ Kf5 21.Rc2 Rg3+ 22.Kf2 a4 23.Rc5+ Ke4 24.Rc4+
  -+  (-2.06)   Depth: 25/39   00:00:01  10860kN
1...Ra5 2.Re2 h5 3.Ke4 h4 4.Rc2 Kg5 5.Kf3 Ra3+ 6.Kg2 a5 7.Rf2 Rg3+ 8.Kh2 Rd3 9.Kg1 a4 10.Kg2 Kg4 11.Kh2 Rh3+ 12.Kg2 Rg3+ 13.Kh2 Rd3 14.Kg2 Rc3 15.Kh2 Re3 16.Kg2 Rg3+
  -+  (-2.14)   Depth: 26/43   00:00:02  17927kN
1...Ra5 2.Re2 h5 3.Ke4 h4 4.Rc2 Kg5 5.Kf3 Ra3+ 6.Kg2 a5 7.Rf2 Rg3+ 8.Kh2 Rd3 9.Kg1 a4 10.Kg2 Kg4 11.Kh2 Rh3+ 12.Kg2 Rg3+ 13.Kh2 Rd3 14.Kg2 Rc3 15.Kh2 Re3 16.Kg2 Rg3+
  -+  (-2.14)   Depth: 27/43   00:00:02  21510kN
1...Ra5 2.Re2 h5 3.Ke4 h4 4.Rc2 Kg5 5.Kf3 Ra3+ 6.Kg2 a5 7.Rf2 Rg3+ 8.Kh2 Rd3 9.Kg1 a4 10.Kg2 Kg4 11.Kh2 Rh3+ 12.Kg2 Rg3+ 13.Kh2 Rd3 14.Kg2 Rc3 15.Kh2 Re3 16.Kg2 Kg5 17.Rb2 Rg3+ 18.Kf2 Ra3 19.Rb5+ Kg4 20.Rb4+ Kf5 21.Rxh4 Rxa2+ 22.Ke3 a3 23.Rh5+ Kg6 24.Rh1 Rb2 25.Ra1 a2 26.Kd4 Kf6
  -+  (-2.14)   Depth: 28/43   00:00:03  34084kN
1...Ra5 2.Re2 h5 3.Ke4 h4 4.Rc2 Kg5 5.Kf3 Ra3+ 6.Kg2 Kf5 7.Rf2+ Ke4 8.Re2+ Kd4 9.Rd2+ Rd3 10.Rf2 Kc3 11.Rf4 Rd2+ 12.Kh3 Rxa2 13.Kxh4 a5 14.Kg5 a4
  -+  (-2.22 !)   Depth: 29/43   00:00:04  44875kN
1...Ra5 2.Re2 h5 3.Ke4 h4 4.Rc2 Kg5 5.Kf3 Ra3+ 6.Kg2 Kf5 7.Rf2+ Ke4 8.Re2+ Kd4 9.Rf2 Ke5 10.Re2+ Kf4 11.Rf2+ Kg4 12.Kh2 a5 13.Kg2 Rg3+ 14.Kh2 a4 15.Rf8 Rd3 16.Rf2 Rh3+ 17.Kg2 Rg3+ 18.Kh2 Rd3 19.Kg2 Rc3 20.Kh2 Re3 21.Kg2 Kg5 22.Rb2 Rg3+
  -+  (-2.14)   Depth: 29/52   00:00:06  58900kN
1...Ra5 2.Re2 h5 3.Ke4 h4 4.Rc2 Kg5 5.Kf3 Ra3+ 6.Kg2 Kf5 7.Rf2+ Ke4 8.Re2+ Re3 9.Rc2 Kd3 10.Rc8 a5 11.Rd8+ Kc3 12.Rc8+ Kb2 13.Rc5 Ra3 14.Rh5 h3+ 15.Kh2 Rxa2 16.Kxh3 a4 17.Kg3 a3
  -+  (-2.22 !)   Depth: 30/52   00:00:07  66901kN
1...Ra5 2.Re2 h5 3.Ke4 h4 4.Rc2 Kg5 5.Rg2+ Kh5 6.Kf4 Ra4+ 7.Kf3 h3 8.Rc2 Kh4 9.Re2 Ra3+ 10.Kf2 Kg4 11.Kg1 Kg3 12.Rg2+ hxg2
  -+  (-2.06 --)   Depth: 30/52   00:00:11  103mN
1...Ra5 2.Re2 h5 3.Ke4 h4
  -+  (-2.30 !)   Depth: 30/52   00:00:12  116mN
1...Ra5 2.Re2 h5 3.Ke4 h4 4.Kf4 Ra4+ 5.Kf3 Kf5 6.Rd2 Ra3+ 7.Kg2 Ke6 8.Kh2 a5 9.Re2+ Kf5 10.Kg2 Rg3+ 11.Kf2 Rc3 12.Rd2 Ke5 13.Re2+ Kd4 14.Kg2 Kd5 15.Kg1 a4 16.Kg2 Rd3 17.Rc2 Kd4 18.Re2 a3 19.Rf2 Kd5 20.Rc2 Rg3+ 21.Kf2
  -+  (-2.30)   Depth: 30/52   00:00:13  118mN
1...Ra5 2.Re2 h5 3.Ke4 h4 4.Kf4 Ra4+ 5.Kf3 Kf5 6.Rd2 Ra3+ 7.Kg2 Ke6 8.Kh2 a5 9.Re2+ Kf5 10.Kg2 Rg3+ 11.Kf2 Rc3 12.Rd2 Ke5 13.Re2+ Kd4 14.Kg2 Kd5 15.Kg1 a4 16.Kg2 Rd3 17.Rc2 Ke5 18.Re2+ Kd5
  -+  (-2.30)   Depth: 31/52   00:00:14  128mN
1...Ra5 2.Re2 h5 3.Ke4 h4 4.Kf4 Ra4+ 5.Kf3 Kf5 6.Rd2 Ra3+ 7.Kg2 Ke6 8.Kh2 a5 9.Re2+ Kf5 10.Kg2 Rg3+ 11.Kf2 Rc3 12.Rd2 Ke5 13.Re2+ Kd4 14.Kg2 Kd5 15.Kg1 a4 16.Kg2 Rd3 17.Rc2 Ke5 18.Re2+ Kf5 19.Rb2 Ra3 20.Re2 Rc3 21.Rb2
  -+  (-2.30)   Depth: 32/52   00:00:16  148mN
1...Ra5 2.Re2 h5 3.Ke4 h4 4.Kf4 Ra4+ 5.Kf3 Kf5 6.Rd2 Ra3+ 7.Kg2 Ke6 8.Kh2 a5 9.Re2+ Kf5 10.Kg2 Rg3+ 11.Kf2 Rc3 12.Rd2 Ke5 13.Re2+ Kd4 14.Kg2 Kd5 15.Kg1 a4 16.Kg2 Rd3 17.Rc2 Ke5 18.Re2+ Kd4 19.Rc2
  -+  (-2.30)   Depth: 33/52   00:00:19  177mN
1...Rd5+ 2.Kc4 Rf5 3.Kd4 Ra5 4.Re6+ Kg5 5.Kc4 Rxa2 6.Re5+ Kg4 7.Re4+ Kf5
  -+  (-2.38 !)   Depth: 34/52   00:00:55  494mN
1...Rd5+ 2.Kc4 Rf5 3.Re2 h5 4.Kd3 Kg5 5.Rg2+ Kf4 6.Ke2 Ra5 7.Kf1 Ra3 8.Rh2 Kg4 9.Kg1 Rg3+ 10.Kh1 Rh3 11.Rxh3 Kxh3
  -+  (-2.46 !)   Depth: 34/59   00:01:24  748mN
1...Rd5+ 2.Kc4 Rf5 3.Re2 h5 4.Kd3 Kg5 5.Rg2+ Kf4 6.Ke2 Ra5 7.Kf1 Ra3 8.Rh2 Kg4 9.Rd2 h4 10.Kg2 Kf5 11.Rf2+ Ke5 12.Re2+ Kd5 13.Rd2+ Ke4 14.Re2+ Re3 15.Rc2 Kd3 16.Rc8 a5 17.Rd8+ Kc3 18.Kf2 Rd3 19.Rc8+ Kb2 20.Rc5 Ra3 21.Rb5+ Kc3 22.Rc5+ Kd4 23.Rc2 Ke5 24.Rd2 Ke6 25.Re2+ Kd5 26.Kg1
  -+  (-2.46)   Depth: 34/59   00:01:35  852mN
Parent - By Dhanish (***) Date 2012-03-27 21:02

> Analysis by Stockfish 120305M


Which is this version of Stockfish? Does it access tablebases? Unfortunately, Stockfish may give evaluations as high as +15.00, only to realize a draw after a few more moves.
- - By Dhanish (***) Date 2012-03-19 16:43
Here is analysis of one more position, which took less than two hours.
Timman-Karpov,1988. The only move to win Rd5+ is found, while Karpov played the drawing Kg5, see http://www.chessgames.com/perl/chessgame?gid=1067025! The mistake was reported by Yakov Konoval earlier, who also discovered this by creating a 7 man tablebase.
Parent - - By Dhanish (***) Date 2012-03-20 02:35
For comparison, I tried IdeA overnight with short time and depth. It could not distinguish between Rd5 and Ra5, in spite of having six men bases. The difficulty is that the Aquarium Engine interface is too slow, only about 10 positions per second get added. Perhaps if this can be improved, making tablebase trees in Aquarium will prove practical.
Parent - - By Master Om (Bronze) Date 2012-03-20 08:09

>It could not distinguish between Rd5 and Ra5, in spite of having six men bases.


What engines you used and how many ? Did all accesed the 6 men TBs ?

Another problem also is it does not understand repeating moves easily. I faced that problem  in my game with Armageddonx in WBCCC12.
Parent - - By Dhanish (***) Date 2012-03-20 12:45

> What engines you used and how many ?


Only one engine, viz Houdini 2.0b.

Today I tried higher time and depth, and IdeA preferred Rd5, though it couldn't establish it as winning.
Parent - - By Master Om (Bronze) Date 2012-03-20 13:20
U set 120s and depth 26 , what was the max time in previous and this case ?
Parent - - By Dhanish (***) Date 2012-03-20 16:47

> max time in previous and this case


In first case no Max time, in second case, Max time 1200 s.

Now I got ambitious and tried FinalGen on the rook ending Hollis-Florian. It started calculating, but then I noticed the estimated time and space requirements, and stopped the run. So, it is not a panacea for all endings!
Parent - By keoki010 (Silver) Date 2012-03-20 19:24

> Now I got ambitious and tried FinalGen on the rook ending Hollis-Florian. It started calculating, but then I noticed the estimated time and space requirements, and stopped the run. So, it is not a panacea for all endings!


:cry: I can see you had better check those every time you try a run or else :cry::cry::cry::cry::fat::mad:
Parent - - By Banned for Life (Gold) Date 2012-03-20 19:56
Actually, given the complexity and number of men in this position, getting a definitive solution in a little over 7 months isn't really too bad!
Parent - - By keoki010 (Silver) Date 2012-03-20 20:04 Edited 2012-03-20 20:07
If you have 24 TB's of storage space :lol:, else chaos :twisted::wink:
[edit] maybe on second look that's 24 TB (tablebases) and 117 GB on the storage.
[edit2] on third it's looking like 24TB storage and 117GB storage. :confused:
Parent - - By Banned for Life (Gold) Date 2012-03-20 20:15
24TB is about $2K. This should be no problem for someone who can allocate a computer to this task full time for 7 months! :grin:
Parent - - By Geomusic (*****) Date 2012-03-22 07:28 Edited 2012-03-22 07:35
24 TB of space and thats 117 GB of memory required. so 2k in harddrives and probably another 2k in memory. But seeing a 10 man TB solved would be quite impressive. I wonder how long its longest mating sequences would be theoretically?
Parent - By Labyrinth (*****) Date 2012-03-22 08:36

>I wonder how long its longest mating sequences would be theoretically?


No kidding. Even more incomprehensible god's chess :-P

Yet to see anyone explain the microscopic progress made by one side in those 500 move mates.
Parent - - By Barnard (Bronze) Date 2012-03-23 03:09
Hi Alan

i tried another similar position with 2 more pawns,but with knight,not rooks (less movements,so i think less moves generated)...

my surprise was when i saw the estimated hard drive needed:near 600 Tb!

if you just change some pawns and you block it,of course,hard drive needs slow too fast,even with the same pices

p.s.just curios,there is a machine that can manage 600 Tb of hard drives?(i mean a comercial machine,not a 'special-built-machine' that just a billionaire can buy)
Parent - - By Banned for Life (Gold) Date 2012-03-23 03:25
Actually, I was impressed that the requirements to solve your 12 man position weren't even more outrageous. 600 TB of disc is certainly manageable and if you had the dough, you could lease this storage from a data center without a problem.
Parent - - By Barnard (Bronze) Date 2012-03-23 03:39
well,maybe i can lease that storage,but if you look at the picture,it isnt showed the time required to build the database,maybe the program was still calculating it when showed the hard drive needed
now it showed me that in my computer,i will need...near 72 years! :yell: i deleted it,but if i dont remember wrong,it showed me about 630.000 hours :eek:
Parent - - By Banned for Life (Gold) Date 2012-03-23 05:21
i will need...near 72 years! :yell:

OK, but wait ten years and computers will have gained enough speed so it will only take about 2 years! :lol:
Parent - By Barnard (Bronze) Date 2012-03-23 06:12
You are right Alan,but if we want to analyze a running endgame (just an example),we cant hold the endgame for 10 years...

but that is an increddible advance generating tablebases,and that using just 1 core...if we add MP (and of course,one hard drive for each core,and of course,one motherboard able to 'play' with that),that 72 years,can be just 9 years with a 8 core machine today...so in your example,will take only 1/8 of your 2 years time,so we will be able (pure speculation),to generate that tablebase in just 3 months :eek:

and that speculating with a maximmum of 8 cores,if FinalGen can reach more cores,even faster...
Parent - - By Uly (Gold) Date 2012-03-20 20:16

> a definitive solution in a little over 7 months isn't really too bad!


But what if it uses the estimated time remaining of Windows?

Parent - By Banned for Life (Gold) Date 2012-03-20 20:18
:yell:
Parent - By keoki010 (Silver) Date 2012-03-20 20:36
:lol::grin::yell::wink::cry: <lmaorofl>
Parent - By Geomusic (*****) Date 2012-03-22 07:26
well ya a 10 man tablebase! LOL
Parent - By Geomusic (*****) Date 2012-03-22 07:37
btw :wink:try the search for draw mode for black instead it won't take as long. It might only be 2 months of calculating !
Parent - - By Barnard (Bronze) Date 2012-03-23 03:01
Hi

see if i was surprised when i saw the needed hard disk to generate the tablebases on that endgame Barcza-Sanchez
Parent - - By Geomusic (*****) Date 2012-03-23 05:48 Edited 2012-03-23 06:01
:P for that one 12 man...593 TB and almost 1 TB of ram good lord.   :) Lukas could your cluster handle this? I wonder if one could upload the program into a cloud like here offering up to 1000 TB's of storage. LOL http://www.drivehq.com/premium/enterprise.aspx I bet they wouldn't enjoy you executing a program there requiring that much memory LOL.
Parent - - By Barnard (Bronze) Date 2012-03-23 06:18 Edited 2012-03-23 06:20
Hi

not,it isnt 593 Tb of hard drive storage and 1Tb ( 986 Mb) of RAM;the first and the second number (986 Gb in that case) both means the storage needed of hard drive to store the tablebases...so is 593 Tb and 986 Gb of hard drive needed to store the database,not the RAM needed

just an example: 130 Gb 245 Mb ----> means 130 Gb and 245 Mb of space needed in the hard drive
Parent - - By Geomusic (*****) Date 2012-03-23 07:16
why are there 2 numbers to represent harddrive space?
Parent - - By Pia (****) Date 2012-03-23 11:52 Edited 2012-03-23 12:03
Me tink its HDD and RAM
"For example, many 7-man tablebases [Nalimov] can be generated using 4 GB of RAM [and up to 14]" and here is 12...
Parent - By Barnard (Bronze) Date 2012-03-23 23:52
No Pia,is hard drive used,the 2 numbers...

download the program,install it,and choose 'english' as default language,and you will see that that 2 numbers are only to tell the hard drive needed

in fact,in the web page of the author,is specified that the programs only need 1 Gb of RAM to run
Parent - By Barnard (Bronze) Date 2012-03-24 01:01
Pia,look at the system requeriments,from the author's web page:

>System requirements


>- Pentium 4, equivalent or better, AMD Athlon 64 or better. FinalGen supports all INTEL processors released after 2000 and AMD processors introduced after 2003.


>- 1 Gigabyte of RAM.


>- Windows 2000/XP/Vista/7.


>- The required hard disk space depends on the characteristics of each particular position. A reasonable amount of disk space is 80GB (or 30GB if you use a compressed folder).


just 1 Gigabyte of RAM is needed to build whatever database,if the program is able to build that database,of course (no more than 1 pice for each side,and from what i tried,if there isnt exceptions,no more than 14 pieces in total)
Parent - - By Barnard (Bronze) Date 2012-03-23 23:50 Edited 2012-03-23 23:52
because in big database that use 1 Tb(just an example),if the second number never is given to you,you wont know if is 1 Tb or 1 Tb AND 999 GB...and is very different;in fact one number is the double of the other

so there is 2 numbers,the first,tells you the Tb (or Gb) and the second the Gb (or the Mb if the first told you Gb)

example 1 Tb and 345 Gb

              24 Gb and 357 Mb

as you can see,the second number is to make more exact the harddrive used
Parent - - By Pia (****) Date 2012-03-24 06:43
Ok, now I tink you're right. Have tested myself, program reported "2 GB, 544 MB" and actually wrote 2.541 GB into folder.
With only visible 3 MB RAM used...
Parent - By Barnard (Bronze) Date 2012-03-24 15:22
:smile:

and yes,that program uses a really small amount of RAM
Parent - By Geomusic (*****) Date 2012-04-26 06:48
well duh its an 11 man tablebase. LOL
Parent - - By ernest (****) Date 2012-03-21 00:02

> Today I tried higher time and depth, and IdeA preferred Rd5


My Houdini 2.0c, with short time infinite analysis... 
1. -/+  (-0.77): 50...Rd5+ 51.Kc4 Rf5 52.Re6+ Rf6 53.Re2 h5 54.Kd4 h4 55.Rh2 Kg5 56.Ke3 Kg4 57.Rg2+ Kh3 58.Rf2 Rg6 59.Kf4 a5 60.Kf5 Rg7 61.Rd2 Kg3 62.Rd3+ Kh2 63.a4 Rg2 64.Ra3 h3 65.Kf4 Rg7 66.Rb3 Kg2 67.Rb2+ Kh1 68.Rb1+ Rg1 69.Rb5
2. =/+  (-0.61): 50...Kf6 51.Rh7 Rh5 52.Rh8 Rh2 53.Kc3 Kg7 54.Rd8 Rxa2 55.Rd7+ Kg6 56.Rd6+ Kg5 57.Rd5+ Kg4 58.Rd4+ Kg3 59.Rd3+ Kh4 60.Rd4+ Kg3 61.Rd3+
3. =/+  (-0.59): 50...Ra5 51.Re2 Kg5 52.Ke4 Kg4 53.Rg2+ Kh3 54.Rg1 h5 55.Kf4 Ra4+ 56.Kf3 Kh4 57.Rg2 Ra3+ 58.Kf4 Kh3 59.Rg1 Ra4+ 60.Kf3


who needs IdeA?... :lol:
Up Topic The Rybka Lounge / Computer Chess / FinalGen
1 2 3 4 5 Previous Next  

Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill