Sting-sf-8

Discussion of anything and everything relating to chess playing software and machines.

Moderator: Ras

carldaman
Posts: 2287
Joined: Sat Jun 02, 2012 2:13 am

Re: Sting-sf-8

Post by carldaman »

Paloma wrote:
lech wrote:
giovanni wrote:
lech wrote:The new version of this project is a next step in my searching for a quasi-intelligent chess engine.

I don't know if you may see the correct solution of each difficult position, but I hope you are much closer to get it now.

Sting-sf-8 is still based on an old version of Stockfish-211.

download: http://www.mediafire.com/file/00eyltvsm ... g-sf-8.zip
Nice, thank you.
When you say "quasi-intelligent", means that you encoded in it is some sort of learning capability?
I see that many of you still like my idea. :D

Sting learns only by my deep understanding of possible problems with difficult positions for standard engines.

BTW.
Note that now Sting starts with THREADS = 1.
Hi Marek,

you wrote last summer (July 18th) here:
http://www.talkchess.com/forum/viewtopi ... ht=#679563

> It is probably the last version based on old Stockfish 2.1.1. I will try to move my ideas to the newer 3.0 version. [/url]

Will this lead to an advantage over version 2.1.1 at all?

Is it not better right away go to SF5 or SF6 ?

But I guess the code of SF5 / 6 is completely different, so the transfer of your ideas is now technically not possible without starting from the beginning.
No matter, thanks for your Work on Sting and the constant improvement!
SF3 is not a great version of SF. It dates back to a time just before the testing framework really took off, and significant improvements came after it.
I agree that SF6 would be a much better choice, but it's up to Marek - we appreciate his efforts. :)
lech
Posts: 1175
Joined: Sun Feb 14, 2010 10:02 pm

Re: Sting-sf-8

Post by lech »

carldaman wrote:
Paloma wrote: Hi Marek,

you wrote last summer (July 18th) here:
http://www.talkchess.com/forum/viewtopi ... ht=#679563

> It is probably the last version based on old Stockfish 2.1.1. I will try to move my ideas to the newer 3.0 version. [/url]

Will this lead to an advantage over version 2.1.1 at all?

Is it not better right away go to SF5 or SF6 ?

But I guess the code of SF5 / 6 is completely different, so the transfer of your ideas is now technically not possible without starting from the beginning.
No matter, thanks for your Work on Sting and the constant improvement!
SF3 is not a great version of SF. It dates back to a time just before the testing framework really took off, and significant improvements came after it.
I agree that SF6 would be a much better choice, but it's up to Marek - we appreciate his efforts. :)
Sting is a tough project and needs much more time than I am able to sacrifice to do it in a perfect way.
Of course it would be good to move my project to a higher version of Stockfish and to see what it gives now.
Indeed, I have thought about the version of Stockfish 3 first.

At this moment, you can try to solve each difficult position. But it doesn't mean that I recommend you to bay a much better computer to check it. :)
Maybe, I can't be friendly, but let me be useful.
Jamal Bubker
Posts: 328
Joined: Mon May 24, 2010 4:32 pm

Re: Sting-sf-8

Post by Jamal Bubker »

Thank you Marek !
Paloma
Posts: 1227
Joined: Thu Dec 25, 2008 9:07 pm
Full name: Herbert L

Re: Bug in Sting-sf-8

Post by Paloma »

Hi Marek,

on my System i7-6700 - 4 cores 8 threads, Win 7 x64, Sting shows during creating a Uci-Engine >Threads=8<

When i start Sting now, it shows in ChessBase 14 >4 CPU< and around 9500 kN/s.
Change it to 1 CPU it shows still 9500 kN/s - Task Manager shows 61% ! CPU-Utilization.

When i set >Threads=1< in Uci-File and starts Sting now, it shows >5 CPU< 2500 kN/s and 12% ! CPU-Utilization.
Increase to 4 or 8 Threads in CB-GUI shows still 2500 kN/s and 12% ! CPU-Utilization.

All other Engines have not this behavior.
lech
Posts: 1175
Joined: Sun Feb 14, 2010 10:02 pm

Re: Bug in Sting-sf-8

Post by lech »

Paloma wrote:Hi Marek,

on my System i7-6700 - 4 cores 8 threads, Win 7 x64, Sting shows during creating a Uci-Engine >Threads=8<

When i start Sting now, it shows in ChessBase 14 >4 CPU< and around 9500 kN/s.
Change it to 1 CPU it shows still 9500 kN/s - Task Manager shows 61% ! CPU-Utilization.

When i set >Threads=1< in Uci-File and starts Sting now, it shows >5 CPU< 2500 kN/s and 12% ! CPU-Utilization.
Increase to 4 or 8 Threads in CB-GUI shows still 2500 kN/s and 12% ! CPU-Utilization.

All other Engines have not this behavior.
Herbert, I can't understend you. Sting Sf 8 should start with THREADS = 1. I don't use Chessbase and I can't help you.
Maybe, I can't be friendly, but let me be useful.
Dann Corbit
Posts: 12870
Joined: Wed Mar 08, 2006 8:57 pm
Location: Redmond, WA USA

Re: Bug in Sting-sf-8

Post by Dann Corbit »

The thread count and NPS seem to work fine with Arena.
I only tested it there.

Live run:

Code: Select all

FEN: 4r1k1/1r1np3/1pqp1ppB/p7/2b1P1PQ/2P2P2/P3B2R/3R2K1 w - - 0 1 

Sting-sf-80:
   2	00:00	       1,130	2,539	-0.28	Bd3 Qc5+ Kf1 Bxd3+ Rxd3
   3	00:00	       2,580	5,771	+0.36	Bxc4+ Qxc4 Be3 Kf7 Qh7+ Kf8 Qxg6 Qxc3
   4	00:00	       3,995	8,917	+2.90	Bxc4+ Qxc4 Be3 Rc8 Qh6 Qxc3
   5-	00:00	       4,418	9,839	+2.77	Bxc4+
   5-	00:00	       4,601	10,247	+2.65	Bxc4+
   5-	00:00	       5,107	11,348	+2.20	Bxc4+
   5	00:00	       5,585	12,411	+1.85	Bxc4+ Qxc4 Be3 Nf8 Rd4 Qxc3
   6-	00:00	       7,704	16,857	+1.73	Bxc4+
   6-	00:00	       8,478	18,551	+1.61	Bxc4+
   6+	00:00	       9,136	19,947	+1.97	Bxc4+
   6	00:00	       9,820	21,441	+1.53	Bxc4+ Qxc4 Be3 Nf8 Qh8+ Kf7 Bh6 Qc5+ Kg2
   7+	00:00	      11,026	23,762	+1.65	Bxc4+
   7+	00:00	      11,273	24,243	+1.77	Bxc4+
   7+	00:00	      12,862	27,600	+3.27	Bxc4+
   7	00:00	      15,275	32,638	+3.11	Bxc4+ Qxc4 Be3 Nf8 Qh8+ Kf7 Bh6 Qxc3 Qg7+ Ke6 Rf2
   8+	00:00	      22,718	47,928	+3.23	Be3
   8+	00:00	      23,048	48,522	+3.35	Be3
   8+	00:00	      23,436	49,338	+3.55	Be3
   8	00:00	      24,497	51,464	+3.63	Be3 Nf8 Qh8+ Kf7 Bh6 Qc5+ Rd4 Bxe2 Qg7+ Ke6 Rxe2 Qxc3
   9-	00:00	      28,809	60,144	+3.51	Be3
   9-	00:00	      29,956	62,538	+3.39	Be3
   9-	00:00	      86,635	158,963	+2.53	Be3
   9	00:00	     274,600	465,423	+2.14	Be3 Nf8 Qh8+ Kf7 Bh6 Rc8 Qg7+ Ke6 g5 fxg5 f4 Bxe2 Rxe2 Kd7 Bxg5 Qc5+ Kg2
  10-	00:00	     322,796	537,098	+2.02	Be3
  10-	00:00	     352,590	579,917	+1.89	Be3
  10+	00:00	     373,992	608,117	+2.26	Be3
  10	00:00	     374,825	608,482	+2.34	Be3 Nf8 Qh8+ Kf7 Bh6 Rc8 Qg7+ Ke6 Bf4 Bxe2 Rxe2 g5 Bg3 Kd7 Rxd6+ Qxd6 Bxd6 Kxd6
  11	00:00	     480,540	757,949	+2.34	Be3 Nf8 Qh8+ Kf7 Bh6 Rc8 Qg7+ Ke6 Bf4 Bxe2 Rxe2 g5 Bg3 Kd7 Rxd6+ Qxd6 Bxd6 Kxd6
  12+	00:00	     694,923	988,510	+2.46	Be3
  12	00:00	     727,831	1,022,234	+2.34	Be3 Nf8 Qh8+ Kf7 Bh6 Rc8 Qg7+ Ke6 Bf4 Bxe2 Rxe2 g5 Bg3 Kd7 Rxd6+ Qxd6 Bxd6 Kxd6 Qg8
  13+	00:00	     767,495	1,065,965	+2.42	Be3
  13+	00:00	     775,271	1,072,297	+2.50	Be3
  13+	00:00	     787,869	1,083,726	+2.62	Be3
  13	00:00	     867,854	1,161,785	+2.62	Be3 Nf8 Qh8+ Kf7 Bh6 Rc8 Qg7+ Ke6 Bxc4+ Qxc4 Rhd2 Qc5+ Kg2 Kd7 Bf4 Kc6 Bxd6 Ne6 Qf7
  14-	00:00	   1,578,819	1,866,216	+2.50	Be3
  14-	00:00	   1,823,783	2,046,894	+2.38	Be3
  14-	00:01	   2,594,375	2,596,971	+1.92	Be3
  14	00:01	   3,010,541	2,832,117	+1.85	Be3 Nf8 Qh8+ Kf7 Bh6 Rc8 Qg7+ Ke6 Bxc4+ Qxc4 Bf4 Kd7 Rxd6+ Ke8 Rd5 Qxc3 Kf2 e6 Qxb7 exd5 exd5 Rc4 Qxb6
  15	00:01	   4,721,565	3,700,286	+1.89	Be3 Nf8 Qh8+ Kf7 Bh6 Rc8 Qg7+ Ke6 Bxc4+ Qxc4 Bf4 Kd7 Rxd6+ Ke8 Rd5 Qxc3 Kf2 Qb2+ Kg3 Qc3 Rhd2 e6 Qxb7 exd5 Bd6 Nd7 Rxd5
  16	00:01	   6,578,745	4,415,265	+1.89	Be3 Nf8 Qh8+ Kf7 Bh6 Rc8 Qg7+ Ke6 Bxc4+ Qxc4 Bf4 Kd7 Rxd6+ Ke8 Rd5 Qxc3 Kf2 Qb2+ Kg3 Qc3 Rhd2 e6 Qxb7 exd5 Bd6 Nd7 Rxd5
  17-	00:02	  11,377,746	5,512,473	+1.81	Be3
  17-	00:02	  12,469,560	5,619,450	+1.73	Be3
  17-	00:02	  17,467,984	6,142,047	+1.16	Be3
  17	00:03	  20,176,527	6,374,890	+1.13	Be3 Nf8 Qh8+ Kf7 Bh6 Qc5+ Kf1 Bxe2+ Kxe2 Qe5 Rg2 f5 Qxe5 dxe5 gxf5 gxf5 Rg5 fxe4 Rf5+ Kg6 Rxf8 exf3+ Kxf3 Rxf8+ Bxf8
  18-	00:04	  27,970,672	6,777,482	+1.01	Be3
  18-	00:04	  31,012,124	6,877,827	+0.88	Be3
  18+	00:04	  33,140,744	6,875,673	+1.52	Be3
  18	00:04	  34,188,023	6,885,805	+1.05	Be3 Nf8 Qh8+ Kf7 Bh6 Qc5+ Kf1 Bxe2+ Kxe2 Qe5 Rg2 f5 Qxe5 dxe5 gxf5 gxf5 exf5 Rc8 Rg7+ Kf6 Rd6+ exd6 Rxb7 Kxf5 Rf7+ Ke6 Rxf8 Rxc3
  19-	00:05	  42,213,834	7,105,509	+0.92	Be3
  19-	00:06	  46,800,084	7,215,554	+0.80	Be3
  19	00:07	  53,583,732	7,387,802	+0.84	Be3 Nf8 Qh8+ Kf7 Bh6 Qc5+ Kf1 Bxe2+ Kxe2 Qe5 Rg2 f5 Qxe5 dxe5 gxf5 gxf5 exf5 Nd7 Rdg1 Rh8 Rg7+ Ke8 Be3 Rc7 Kd3 Nc5+ Bxc5 Rxc5
  20-	00:08	  61,876,248	7,488,351	+0.72	Be3
  20+	00:10	  82,591,171	7,833,744	+0.96	Be3
  20+	00:10	  84,690,483	7,813,495	+1.83	Be3
  20	00:12	 102,560,084	7,970,165	+1.37	Be3 Nf8 Bxc4+ Qxc4 g5 Qe6 Qh8+ Kf7 Bd4 Reb8 gxf6 exf6 Rg2 Ke7 Qg7+ Ke8 Qxf6 Rf7 Qxe6+ Nxe6 Rxg6 Ke7 Kf2 Rbf8 Rg3 Nxd4 cxd4 Rh7 Rb1 Rh2+ Ke3
  21+	00:14	 112,903,985	8,043,313	+1.49	Be3
  21+	00:14	 113,326,923	8,029,397	+1.61	Be3
  21	00:15	 120,541,658	8,030,221	+1.49	Be3 Nf8 Bxc4+ Qxc4 g5 fxg5 Qh8+ Kf7 Rd5 Rc8 Bd4 Qd3 Qg7+ Ke8 Rh8 Qxf3 Rf5 Qg4+ Kf1 Kd7 Rfxf8 Rxf8+ Rxf8 Qd1+ Kf2 Qc2+ Ke3 Qc1+ Kf3 Qd1+ Kf2
  22-	00:17	 144,538,216	8,099,193	+1.37	Be3
  22-	00:18	 152,888,666	8,104,785	+1.25	Be3
  22-	00:21	 176,329,210	8,175,880	+1.07	Be3
  22	00:22	 180,414,290	8,147,321	+1.17	Be3 Nf8 Bxc4+ Qxc4 g5 fxg5 Qh8+ Kf7 Rd5 Rc8 Bd4 Ke8 Qg7 Qd3 Rh8 Qxf3 Rf5 Qg4+ Kf1 Kd7 Rfxf8 Rxf8+ Qxf8 Qf4+ Bf2 Qxf8 Rxf8 b5 Rg8 b4 cxb4 Rxb4 Rxg6 Rb2 Rxg5 Rxa2
  23-	00:34	 284,717,749	8,318,999	+1.05	Be3
  23+	00:34	 288,013,877	8,319,531	+1.29	Be3
  23	00:35	 294,748,705	8,304,417	+1.05	Be3 Nf8 Bxc4+ Qxc4 g5 fxg5 Qh8+ Kf7 Rd5 Rc8 Bd4 Ke8 Qg7 Qd3 Rh8 Qxf3 Rf5 Qg4+ Kf1 Kd7 Rfxf8 Rxf8+ Qxf8 Qf4+ Kg1 Qxf8 Rxf8 b5 Rg8 g4 Rxg6
  24+	00:36	 302,209,794	8,323,962	+1.17	Be3
  24	00:42	 351,465,859	8,352,325	+1.05	Be3 Nf8 Bxc4+ Qxc4 g5 fxg5 Qh8+ Kf7 Rd5 Rc8 Bd4 Ke8 Qg7 Qd3 Rh8 Qxf3 Rf5 Qg4+ Kf1 Kd7 Rfxf8 Rxf8+ Qxf8 Qf4+ Bf2 Qxf8 Rxf8 Rc7 Rf3 Rc4 Bxb6 Rxe4 Rg3 Ra4 Rxg5 Rxa2 Rxg6
  25+	00:43	 361,590,664	8,366,087	+1.13	Be3
  25	00:51	 432,849,316	8,435,964	+1.13	Be3 Nf8 Bxc4+ Qxc4 g5 fxg5 Qh8+ Kf7 Rd5 Rc8 Bd4 Ke8 Qg8 e6 Rh8 Rf7 Rxd6 g4 Rxb6 Rxf3 Qxe6+ Qxe6 Rxe6+ Kd7 Rb6 Rg3+ Kh2 Rf3
  26+	00:52	 443,272,007	8,453,742	+1.21	Be3
  26	01:08	 581,754,408	8,556,721	+1.09	Be3 Nf8 Bxc4+ Qxc4 g5 fxg5 Qh8+ Kf7 Rd5 Rc8 Bd4 Ke8 Qg8 e6 Rh8 Rf7 Rxd6 g4 Rxb6 Rxf3 Qxe6+ Qxe6 Rxe6+ Kd7 Rb6 a4 Kg2 Rc6 Rb5 Ne6 Rh7+ Kc8 Be5
  27+	01:18	 684,279,139	8,674,168	+1.17	Be3
  27+	01:20	 702,342,338	8,687,302	+1.25	Be3
  27-	01:55	1,026,096,884	8,863,925	+1.01	Be3
  27	02:39	1,434,354,965	8,985,216	+1.33	Be3 Nf8 Bxc4+ Qxc4 g5 fxg5 Qh8+ Kf7 Rd5 Rc8 Bd4 Ke8 Qg8 e6 Rxd6 Rf7 Rh8 Rxf3 Qxe6+ Qxe6 Rxe6+ Kd7 Rxb6 Rf4 Bg7 Rf7 Be5 Rc6 Rb8 Ne6 Ra8 g4
  28-	03:41	2,004,090,356	9,062,787	+1.21	Be3

Taking ideas is not a vice, it is a virtue. We have another word for this. It is called learning.
But sharing ideas is an even greater virtue. We have another word for this. It is called teaching.
peter
Posts: 3570
Joined: Sat Feb 16, 2008 7:38 am
Full name: Peter Martan

Re: Bug in Sting-sf-8

Post by peter »

Hi Herbert!

All I see in chessbase 12 is that the n/s don't gain so much over 50% CUP- usage as for hyperthreading on, and not much more than 80% is reached in Windows Taskmananger, which was normal for SF 2 as far as I remember. Tried an semi- old Intel X5650 Xeon with 12x2,67GHz.
Yet % of CPU- usage in TM, threads-count and n/s correlate reasonably here for me too
Peter.
lech
Posts: 1175
Joined: Sun Feb 14, 2010 10:02 pm

Re: Bug in Sting-sf-8

Post by lech »

lech wrote:
Paloma wrote:Hi Marek,

on my System i7-6700 - 4 cores 8 threads, Win 7 x64, Sting shows during creating a Uci-Engine >Threads=8<

When i start Sting now, it shows in ChessBase 14 >4 CPU< and around 9500 kN/s.
Change it to 1 CPU it shows still 9500 kN/s - Task Manager shows 61% ! CPU-Utilization.

When i set >Threads=1< in Uci-File and starts Sting now, it shows >5 CPU< 2500 kN/s and 12% ! CPU-Utilization.
Increase to 4 or 8 Threads in CB-GUI shows still 2500 kN/s and 12% ! CPU-Utilization.

All other Engines have not this behavior.
Herbert, I can't understend you. Sting Sf 8 should start with THREADS = 1. I don't use Chessbase and I can't help you.
Indeed it can be my bug. I willl return to the old Stockfish 2.1.1 start THERADS = max. possible threads.
Please don't use Sting SF 8 and wait for version 8.1 soon.
Maybe, I can't be friendly, but let me be useful.
Vinvin
Posts: 5325
Joined: Thu Mar 09, 2006 9:40 am
Full name: Vincent Lejeune

Re: Bug in Sting-sf-8

Post by Vinvin »

lech wrote:I willl return to the old Stockfish 2.1.1 start THERADS = max. possible threads.
May be half the max because of all the hyperthreaded CPUs now.