7 Man Syzygy and SSD

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

Moderators: bob, hgm, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
MikeB
Posts: 3393
Joined: Thu Mar 09, 2006 5:34 am
Location: Pen Argyl, Pennsylvania

Re: 7 Man Syzygy and SSD

Post by MikeB » Wed Jul 10, 2019 2:17 am

Nordlandia wrote:
Tue Jul 09, 2019 3:42 pm
Which of the pawnless tables is good to download ?

The partial syzygy set consists of pawnful tables.

http://tablebase.lichess.ovh/tables/standard/7/
Apparently they do not come up that often , I would start with 4/3 ones that are more or less balanced or perhaps interesting/exotic , KQBNKQR , KQNNKQR, KQBNKQB KRNBKRB etc.

User avatar
Nordlandia
Posts: 2410
Joined: Fri Sep 25, 2015 7:38 pm
Location: Sortland, Norway

Re: 7 Man Syzygy and SSD

Post by Nordlandia » Wed Jul 10, 2019 4:01 am

MikeB wrote:
Wed Jul 10, 2019 2:17 am
Nordlandia wrote:
Tue Jul 09, 2019 3:42 pm
Which of the pawnless tables is good to download ?

The partial syzygy set consists of pawnful tables.

http://tablebase.lichess.ovh/tables/standard/7/
Apparently they do not come up that often , I would start with 4/3 ones that are more or less balanced or perhaps interesting/exotic , KQBNKQR , KQNNKQR, KQBNKQB KRNBKRB etc.
Good suggestion as KQBNvKQB is usually a win, so is KRBNvKRN. I had to remove 6-piece nalimov for more space for my new 970 Evo Plus 2tb (used 1 TB out of 1.81 TB so far). First impression is that 7-piece syzygy give more insight and easily solves endgame studies that 6-pieces were insufficient. For the moment i'm downloading WDL because they can be probed during analysis.

User avatar
Nordlandia
Posts: 2410
Joined: Fri Sep 25, 2015 7:38 pm
Location: Sortland, Norway

Re: 7 Man Syzygy and SSD

Post by Nordlandia » Sun Jul 14, 2019 8:48 am

Playing engine games & probing 7-piece syzygy with just barely enough RAM (16 GB in my case) forces windows to terminate cutechess as RAM get filled to 97% very quickly. Windows either terminate cutechess or reboots, anyone with similar experiences ?. I''ll receive 128 GB ram sticks on monday or tuesday.

User avatar
Nordlandia
Posts: 2410
Joined: Fri Sep 25, 2015 7:38 pm
Location: Sortland, Norway

Re: 7 Man Syzygy and SSD

Post by Nordlandia » Mon Jul 15, 2019 4:23 pm

It turns out 128 GB works just fine on my Asus Deluxe II from 2016.

User avatar
Nordlandia
Posts: 2410
Joined: Fri Sep 25, 2015 7:38 pm
Location: Sortland, Norway

Re: 7 Man Syzygy and SSD

Post by Nordlandia » Wed Jul 17, 2019 12:30 pm

I now have about 1.9 TB of WDL on new M.2 SSD and it work great although i want to ask a simple question.

I have the most important tables i believe now although it can never be perfect without them all, as MikeB hinted few posts back.

For analysis WDL is sufficient. Is DTZ worth storing on HDD.

Spliffjiffer
Posts: 195
Joined: Thu Aug 02, 2012 5:48 pm
Location: Germany

Re: 7 Man Syzygy and SSD

Post by Spliffjiffer » Wed Jul 17, 2019 2:16 pm

as far as i know u can easily store the DTZ on a HDD (maybe even USB id believe) !?
if im wrong on this someone will correct me surely.
1,9 TB is a lot, i have only 342 GB of WDL7 which is pretty much sufficient in most cases in terms of verifying a win/draw :-)
Wahrheiten sind Illusionen von denen wir aber vergessen haben dass sie welche sind.

User avatar
Nordlandia
Posts: 2410
Joined: Fri Sep 25, 2015 7:38 pm
Location: Sortland, Norway

Re: 7 Man Syzygy and SSD

Post by Nordlandia » Wed Jul 17, 2019 3:12 pm

For analysis WDL is fine but for engine games DTZ is better to have than not.

Stockfish and LCZero can work with WDL without DTZ, but it might not find the optimal way to win under engine playing conditions. If you are only interested in analyzing positions with tablebases, you don't need the DTZ to tell you if a position is won or loss (since WDL already does that), but having DTZ will guarantee the winning/bestplay sequence if the engine cannot figure it out from probing WDL alone.

Downloading 7-piece tables takes time.

Zenmastur
Posts: 439
Joined: Sat May 31, 2014 6:28 am

Re: 7 Man Syzygy and SSD

Post by Zenmastur » Tue Jul 30, 2019 7:36 pm

MikeB wrote:
Tue Dec 18, 2018 3:53 pm
The purpose of this post is to help one decide which 7 man syzygy files , along with the nearly 200GB files needed for 6 Man Syzygy to place on the a 1TB SSD drive to get maximum benefit.

I'm just going to put this out there as sort of help post for those that are wondering about the most economical way to use 7 man syzygy. I'm sure others have some pretty good suggestions as well and please feel free to share. First, there is real no need to get all 20 GB or so to benefit from 7 man syzygy. Fromm a practical standpoint , about 6o files (rtbw and rtbz) are all you need to get started , provided you have the all of the 6 man syzygy. Only RTBW are used during search, but the RTBZ files are needed, and all the tables that can be reached through promotion and the corresponding DTZ tables to play out the mate reliably. For this purpose, it assumed you will place the others files required to play out mate on an HDD drive.

So we have an 1 TB SSD drive, we have copied all of the 6 man rtbw and rtbz (recommended) onto this drive and we have roughly 800 GB left. A portfolio of the 7 man rtbw files listed below, will capture nearly 80% of the 7 man endings reached in human play.

Code: Select all

Size	EGTB		Bytes
--------------------------------------
    77G	 KBNPvKRP.rtbw	82,917,193,040
    75G	 KRBPvKRB.rtbw	80,280,358,288
    69G	 KRNPvKRN.rtbw	74,082,453,136
    68G	 KRNPvKRB.rtbw	72,853,318,096
    54G	 KRNPvKRP.rtbw	58,025,199,248
    45G	 KRBPvKRP.rtbw	48,381,043,216
    39G	 KRNPvKBN.rtbw	42,185,842,192
    29G	 KBPPvKRP.rtbw	31,606,770,192
    29G	 KBPPvKBP.rtbw	30,746,474,768
    28G	 KNPPvKRP.rtbw	29,911,219,856
    27G	 KNPPvKNP.rtbw	29,087,503,696
    27G	 KBPPvKNP.rtbw	28,727,898,832
    27G	 KRPPvKRB.rtbw	28,687,360,336
    26G	 KRPPvKRN.rtbw	27,834,688,208
    22G	 KRPPvKRP.rtbw	23,261,437,520
    21G	 KQPPvKQP.rtbw	22,632,171,664
    16G	 KRPPvKQP.rtbw	16,689,932,048
    11G	 KRPPvKBP.rtbw	11,435,368,336
   9.0G	 KNPPvKPP.rtbw	9,689,699,280
   9.0G	 KRPPvKNP.rtbw	9,632,432,848
   7.4G	 KBPPvKPP.rtbw	7,973,639,248
   7.1G	 KPPPvKBP.rtbw	7,584,064,976
   6.2G	 KPPPvKNP.rtbw	6,680,648,720
   4.7G	 KRPPvKPP.rtbw	5,054,627,664
   4.3G	 KPPPvKRP.rtbw	4,645,229,072
   2.1G	 KPPPvKQP.rtbw	2,269,738,128
   2.0G	 KPPPvKPP.rtbw	2,107,379,792
   1.0G	 KRPPPvKR.rtbw	1,075,778,832
   686M	 KRNPvKQP.rtbw	719,440,845
   ------------------------------------
			796,778,912,077
Happy Endgame Analysis!

Hmmm... I did a list of files based on the number of occurrences in a 7M game database. They are sorted, most important first. Importance is given by the size of the files divided by the number of times an endgames is seen in the database.

Code: Select all

                                                        SUM   		Number/
		Pieces	# in DB	  %	  SUM	  %	GB	 Gb	per MB
1	4+3	krppkrp	165,068	2.33%	165,068	26.21%	2.3	2.326	70.962
2	4+3	kpppkpp	60,353	0.85%	225,421	35.79%	4.4	2.107	28.639
3	5+2	krpppkr	20,016	0.28%	245,437	38.97%	5.5	1.076	18.606
4	5+2	kppppkp	3,933	0.06%	249,370	39.59%	5.8	0.337	11.672
5	5+2	kqpppkp	1,934	0.03%	251,304	39.90%	6.0	0.189	10.231
6	4+3	kqppkpp	7,116	0.10%	258,420	41.03%	8.1	2.110	3.373
7	5+2	krpppkp	3,337	0.05%	261,757	41.56%	9.2	1.093	3.054
8	5+2	krpppkn	552	0.01%	262,309	41.65%	9.4	0.189	2.914
9	4+3	kqqqknp	604	0.01%	262,913	41.74%	9.6	0.220	2.742
10	5+2	kbpppkn	1,770	0.03%	264,683	42.03%	10.3	0.664	2.666
11	5+2	krpppkb	676	0.01%	265,359	42.13%	10.6	0.257	2.629
12	4+3	krppkpp	12,457	0.18%	277,816	44.11%	15.6	5.055	2.464
13	5+2	knpppkn	2,322	0.03%	280,138	44.48%	16.6	0.960	2.419
14	4+3	kpppkqp	5,008	0.07%	285,146	45.27%	18.9	2.270	2.206
15	5+2	krrppkr	903	0.01%	286,049	45.42%	19.3	0.469	1.926
16	4+3	kpppkrp	8,882	0.13%	294,931	46.83%	24.0	4.645	1.912
17	4+3	kbppkpp	15,047	0.21%	309,978	49.22%	31.9	7.974	1.887
18	5+2	kbpppkp	2,746	0.04%	312,724	49.65%	33.4	1.492	1.840
19	5+2	krbppkr	3,125	0.04%	315,849	50.15%	35.4	1.930	1.619
20	5+2	kbpppkb	2,812	0.04%	318,661	50.60%	37.2	1.798	1.564
21	4+3	knppkpp	13,315	0.19%	331,976	52.71%	46.8	9.690	1.374
22	5+2	kppppkn	596	0.01%	332,572	52.80%	47.3	0.440	1.355
23	5+2	kqrppkp	744	0.01%	333,316	52.92%	47.8	0.556	1.338
24	5+2	knpppkb	1,883	0.03%	335,199	53.22%	49.3	1.447	1.301
25	4+3	kpppkbp	9,839	0.14%	345,038	54.78%	56.9	7.584	1.297
26	4+3	kpppknp	8,578	0.12%	353,616	56.15%	63.6	6.681	1.284
27	5+2	knpppkp	2,214	0.03%	355,830	56.50%	65.3	1.732	1.278
28	5+2	kqbppkp	745	0.01%	356,575	56.62%	65.9	0.606	1.229
29	5+2	kqpppkr	314	0.00%	356,889	56.67%	66.2	0.276	1.139
30	5+2	krnppkr	2,604	0.04%	359,493	57.08%	68.6	2.405	1.083
31	5+2	kppppkb	587	0.01%	360,080	57.17%	69.2	0.632	0.928
32	4+3	kbppkbp	26,184	0.37%	386,264	61.33%	100.0	30.746	0.852
33	4+3	krppkbp	9,633	0.14%	395,897	62.86%	111.4	11.435	0.842
34	5+2	kqrppkr	569	0.01%	396,466	62.95%	112.1	0.709	0.803
35	5+2	kqnppkp	495	0.01%	396,961	63.03%	112.7	0.634	0.781
36	4+3	kqqppkp	203	0.00%	397,164	63.06%	113.0	0.270	0.753
37	4+3	knppknp	21,387	0.30%	418,551	66.46%	142.1	29.088	0.735
38	4+3	knppkbp	20,993	0.30%	439,544	69.79%	173.0	30.956	0.678
39	4+3	kbppknp	19,074	0.27%	458,618	72.82%	201.8	28.728	0.664
40	4+3	krppknp	6,275	0.09%	464,893	73.81%	211.4	9.632	0.651
41	5+2	kppppkr	1,090	0.02%	465,983	73.99%	213.2	1.763	0.618
42	4+3	kqppkqp	13,494	0.19%	479,477	76.13%	235.8	22.632	0.596
43	5+2	kqpppkq	2,006	0.03%	481,483	76.45%	239.5	3.746	0.536
44	4+3	kqqpkpp	169	0.00%	481,652	76.48%	239.9	0.346	0.488
45	4+3	kqppkrp	2,364	0.03%	484,016	76.85%	245.4	5.507	0.429
46	4+3	kqrpkrp	2,570	0.04%	486,586	77.26%	251.5	6.058	0.424
47	4+3	kbppkrp	13,357	0.19%	499,943	79.38%	283.1	31.607	0.423
48	4+3	kqrpkpp	558	0.01%	500,501	79.47%	284.4	1.322	0.422
49	5+2	krbppkb	342	0.00%	500,843	79.52%	285.2	0.818	0.418
50	4+3	krppkqr	3,095	0.04%	503,938	80.01%	292.9	7.690	0.402
51	4+3	krbppkn	215	0.00%	504,153	80.05%	293.5	0.581	0.370
52	5+2	kppppkq	426	0.01%	504,579	80.12%	294.7	1.199	0.355
53	5+2	kbnppkn	308	0.00%	504,887	80.16%	295.6	0.899	0.343
54	5+2	kbpppkr	2,243	0.03%	507,130	80.52%	302.5	6.897	0.325
55	5+2	kbnppkb	455	0.01%	507,585	80.59%	303.9	1.451	0.314
56	4+3	krnppkn	186	0.00%	507,771	80.62%	304.5	0.627	0.297
57	4+3	knppkrp	8,395	0.12%	516,166	81.96%	334.5	29.911	0.281
58	4+3	kqppkbp	776	0.01%	516,942	82.08%	337.5	3.066	0.253
59	4+3	krnppkb	175	0.00%	517,117	82.11%	338.2	0.696	0.251
60	4+3	krppkqp	4,188	0.06%	521,305	82.77%	354.9	16.690	0.251
61	4+3	kqbpkpp	699	0.01%	522,004	82.88%	357.7	2.790	0.251
62	4+3	krbpkrp	11,999	0.17%	534,003	84.79%	406.1	48.381	0.248
63	5+2	krbppkp	750	0.01%	534,753	84.91%	409.2	3.067	0.245
64	4+3	kpppkqr	254	0.00%	535,007	84.95%	410.2	1.071	0.237
65	4+3	kqppknp	766	0.01%	535,773	85.07%	413.5	3.316	0.231
66	5+2	knpppkr	1,461	0.02%	537,234	85.30%	420.8	7.312	0.200
67	4+3	krrppkp	215	0.00%	537,449	85.33%	422.0	1.103	0.195
68	4+3	kqnpkpp	629	0.01%	538,078	85.43%	425.3	3.344	0.188
69	4+3	krrpkrp	1,653	0.02%	539,731	85.70%	434.7	9.366	0.176
70	4+3	krnpkrp	10,056	0.14%	549,787	87.29%	492.7	58.025	0.173
71	4+3	krppkrb	4,784	0.07%	554,571	88.05%	521.4	28.687	0.167
72	5+2	krnppkp	559	0.01%	555,130	88.14%	524.9	3.508	0.159
73	4+3	krppkrn	4,111	0.06%	559,241	88.79%	552.7	27.835	0.148
74	4+3	krppkrr	1,279	0.02%	560,520	89.00%	564.5	11.768	0.109
75	4+3	knppkqp	1,158	0.02%	561,678	89.18%	575.9	11.438	0.101
76	4+3	kbppkqp	1,090	0.02%	562,768	89.35%	588.1	12.139	0.090
77	4+3	krbpkpp	951	0.01%	563,719	89.51%	599.2	11.152	0.085
78	5+2	kbnppkr	1,206	0.02%	564,925	89.70%	615.4	16.209	0.074
79	4+3	kpppkqn	213	0.00%	565,138	89.73%	618.6	3.205	0.066
80	4+3	kbnpkpp	1,074	0.02%	566,212	89.90%	634.9	16.320	0.066
81	4+3	kpppkqb	199	0.00%	566,411	89.93%	638.0	3.033	0.066
82	5+2	kbnppkp	475	0.01%	566,886	90.01%	645.2	7.263	0.065
83	4+3	krpppkq	694	0.01%	567,580	90.12%	656.4	11.111	0.062
84	4+3	kqbpkbp	571	0.01%	568,151	90.21%	665.5	9.149	0.062
85	4+3	krrpkrr	1,271	0.02%	569,422	90.41%	686.2	20.655	0.062
86	4+3	kqrpkbp	202	0.00%	569,624	90.44%	689.4	3.283	0.062
87	4+3	krnpkpp	773	0.01%	570,397	90.57%	702.4	12.964	0.060
88	4+3	kpppkrb	333	0.00%	570,730	90.62%	708.6	6.160	0.054
89	4+3	kqnpkbp	501	0.01%	571,231	90.70%	718.1	9.551	0.052
90	4+3	krrpkpp	172	0.00%	571,403	90.73%	721.7	3.579	0.048
91	4+3	kqppkrb	451	0.01%	571,854	90.80%	731.3	9.621	0.047
92	4+3	knnpknp	419	0.01%	572,273	90.86%	740.5	9.223	0.045
93	4+3	krrpkbp	284	0.00%	572,557	90.91%	747.3	6.730	0.042
94	4+3	kqbpkrp	499	0.01%	573,056	90.99%	759.5	12.271	0.041
95	4+3	krppkbn	1,144	0.02%	574,200	91.17%	788.5	28.996	0.039
96	4+3	kbbppkb	235	0.00%	574,435	91.21%	794.5	5.991	0.039
97	4+3	kpppkrn	285	0.00%	574,720	91.25%	801.8	7.288	0.039
98	4+3	krbpkrb	3,105	0.04%	577,825	91.75%	882.1	80.280	0.039
99	4+3	kpppkbn	381	0.01%	578,206	91.81%	892.4	10.309	0.037
100	4+3	krbpkbp	970	0.01%	579,176	91.96%	919.4	26.999	0.036
101	4+3	kqnpknp	365	0.01%	579,541	92.02%	929.6	10.210	0.036
102	4+3	kqppkrn	304	0.00%	579,845	92.07%	938.2	8.566	0.035
103	4+3	kqppkqq	194	0.00%	580,039	92.10%	943.7	5.493	0.035
104	4+3	krrpkqp	1,226	0.02%	581,265	92.29%	979.1	35.444	0.035
105	4+3	kbbpkbp	1,131	0.02%	582,396	92.47%	1,012.2	33.116	0.034
106	4+3	kbnpkrp	2,814	0.04%	585,210	92.92%	1,095.1	82.917	0.034
107	4+3	kbppkqr	201	0.00%	585,411	92.95%	1,101.1	5.945	0.034
108	4+3	kqbpknp	302	0.00%	585,713	93.00%	1,110.1	9.010	0.034
109	4+3	kbbpkpp	290	0.00%	586,003	93.04%	1,118.9	8.819	0.033
110	4+3	krbpkrn	2,498	0.04%	588,501	93.44%	1,199.1	80.189	0.031
111	4+3	krnpkrn	2,275	0.03%	590,776	93.80%	1,273.2	74.082	0.031
112	4+3	krnpkrb	2,236	0.03%	593,012	94.16%	1,346.0	72.853	0.031
113	4+3	kbpppkq	238	0.00%	593,250	94.19%	1,353.9	7.843	0.030
114	4+3	knpppkq	257	0.00%	593,507	94.24%	1,362.4	8.506	0.030
115	4+3	kqbppkq	253	0.00%	593,760	94.28%	1,371.0	8.604	0.029
116	4+3	kbppkbb	301	0.00%	594,061	94.32%	1,381.5	10.506	0.029
117	4+3	krrpkrn	435	0.01%	594,496	94.39%	1,396.7	15.234	0.029
118	4+3	kbnpkbp	1,942	0.03%	596,438	94.70%	1,465.5	68.807	0.028
119	4+3	kbppkqb	288	0.00%	596,726	94.75%	1,475.8	10.213	0.028
120	4+3	krrpkrb	537	0.01%	597,263	94.83%	1,495.2	19.427	0.028
121	4+3	kqppkrr	480	0.01%	597,743	94.91%	1,512.7	17.494	0.027
122	4+3	krrpknp	157	0.00%	597,900	94.93%	1,518.7	6.000	0.026
123	4+3	kqrpkqp	774	0.01%	598,674	95.06%	1,548.3	29.589	0.026
124	4+3	kbbppkr	252	0.00%	598,926	95.10%	1,558.1	9.782	0.026
125	4+3	krbpknp	574	0.01%	599,500	95.19%	1,581.0	22.985	0.025
126	4+3	kbppkrb	818	0.01%	600,318	95.32%	1,614.9	33.900	0.024
127	4+3	kqnpkrp	334	0.00%	600,652	95.37%	1,629.1	14.197	0.024
128	4+3	kbppkbn	581	0.01%	601,233	95.46%	1,655.9	26.785	0.022
129	4+3	kbppkqn	251	0.00%	601,484	95.50%	1,667.8	11.844	0.021
130	4+3	knnpkpp	272	0.00%	601,756	95.55%	1,680.8	12.996	0.021
131	4+3	krppknn	251	0.00%	602,007	95.59%	1,693.0	12.229	0.021
132	4+3	kqnppkq	200	0.00%	602,207	95.62%	1,702.7	9.754	0.021
133	4+3	krnpkbp	600	0.01%	602,807	95.71%	1,733.8	31.085	0.019
134	5+2	krrppkq	360	0.01%	603,167	95.77%	1,752.6	18.773	0.019
135	4+3	kbbpkrp	734	0.01%	603,901	95.89%	1,791.6	39.030	0.019
136	4+3	knppknn	179	0.00%	604,080	95.91%	1,801.3	9.668	0.019
137	4+3	kbnpknp	1,183	0.02%	605,263	96.10%	1,869.8	68.483	0.017
138	4+3	knppkbn	457	0.01%	605,720	96.17%	1,896.3	26.524	0.017
139	4+3	kqbpkqp	920	0.01%	606,640	96.32%	1,951.2	54.899	0.017
140	4+3	krppkqb	247	0.00%	606,887	96.36%	1,966.1	14.886	0.017
141	4+3	knnppkr	236	0.00%	607,123	96.40%	1,980.7	14.609	0.016
142	4+3	krbpkrr	611	0.01%	607,734	96.49%	2,018.8	38.089	0.016
143	4+3	krbpkqp	1,122	0.02%	608,856	96.67%	2,090.7	71.950	0.016
144	4+3	knppkqb	185	0.00%	609,041	96.70%	2,102.7	11.919	0.016
145	4+3	knppkrb	497	0.01%	609,538	96.78%	2,136.2	33.530	0.015
146	4+3	kqqpkqp	327	0.00%	609,865	96.83%	2,158.8	22.632	0.014
147	4+3	knppkqn	199	0.00%	610,064	96.86%	2,172.6	13.785	0.014
148	4+3	kqnpkqp	864	0.01%	610,928	97.00%	2,233.9	61.245	0.014
149	4+3	knppkrn	480	0.01%	611,408	97.08%	2,270.0	36.147	0.013
150	4+3	kqppkqb	410	0.01%	611,818	97.14%	2,302.6	32.584	0.013
151	4+3	kqppkqr	318	0.00%	612,136	97.19%	2,329.2	26.648	0.012
152	4+3	krppkbb	197	0.00%	612,333	97.22%	2,345.7	16.508	0.012
153	4+3	knnpkrp	435	0.01%	612,768	97.29%	2,385.2	39.476	0.011
154	4+3	krnpkqp	719	0.01%	613,487	97.41%	2,450.9	65.656	0.011
155	4+3	kqppkqn	355	0.01%	613,842	97.46%	2,484.0	33.123	0.011
156	4+3	kbppkrn	439	0.01%	614,281	97.53%	2,525.3	41.313	0.011
157	4+3	kbnpkbb	231	0.00%	614,512	97.57%	2,548.3	23.033	0.010
158	4+3	krnpkrr	403	0.01%	614,915	97.63%	2,589.2	40.859	0.010
159	4+3	kbnpkqp	274	0.00%	615,189	97.68%	2,618.1	28.904	0.009
160	4+3	krppkqn	167	0.00%	615,356	97.70%	2,636.3	18.162	0.009
161	4+3	kbbpkbn	345	0.00%	615,701	97.76%	2,673.8	37.557	0.009
162	4+3	kbbpknp	335	0.00%	616,036	97.81%	2,710.4	36.600	0.009
163	5+2	krbppkq	399	0.01%	616,435	97.88%	2,754.2	43.818	0.009
164	4+3	kbnpkbn	428	0.01%	616,863	97.94%	2,805.5	51.298	0.008
165	4+3	krbpkbn	267	0.00%	617,130	97.99%	2,838.0	32.434	0.008
166	4+3	knnpkbp	372	0.01%	617,502	98.05%	2,886.5	48.566	0.008
167	4+3	knppkbb	176	0.00%	617,678	98.07%	2,911.7	25.142	0.007
168	4+3	kqrpkqr	507	0.01%	618,185	98.15%	2,986.0	74.314	0.007
169	4+3	krnpkqr	163	0.00%	618,348	98.18%	3,015.8	29.800	0.005
170	4+3	krnppkq	253	0.00%	618,601	98.22%	3,062.8	47.013	0.005
171	4+3	krbpkqr	167	0.00%	618,768	98.25%	3,098.7	35.920	0.005
172	4+3	krnpkbn	179	0.00%	618,947	98.27%	3,140.9	42.186	0.004
173	4+3	krbnkrp	219	0.00%	619,166	98.31%	3,198.0	57.068	0.004
174	4+3	kqnpkqb	181	0.00%	619,347	98.34%	3,277.4	79.436	0.002
175	4+3	kbnpkrb	253	0.00%	619,600	98.38%	3,418.6	141.168	0.002

These number will vary slightly depending on the database of games used. The idea is to be able to handle the largest fraction of 7-man endgames with the given space limitations.

I didn't consider all files. Some of the endgames never occurred in the database and others rarely occurred. If the endgame didn't occur at least once every 50,000 games I left it out.

First % column is % of all games that ended in this endgame. Second % column is the % of all endgames that will be handled if you include this and all files above it. Sum GB is the sum of the sizes of this and all files above it in the list. GB is the size of this file.
Last column is the relative importance of this file (higher is more important).

I did this late last night so there may be some errors in copying the numbers etc. If you see any let me know.

EDIT: Some odd endgames are included because they occur often enough and have small or very-small files. It's more "important" to include them than some more common endings. Therefore they appear higher in the list! Some common endings have HUGE file sizes and therefore appear lower on the list than you might expect. The original list omitted at least one "important" file and included several others that weren't very important.


Regards,

Zenmastur
Last edited by Zenmastur on Tue Jul 30, 2019 7:56 pm, edited 1 time in total.
Only 2 defining forces have ever offered to die for you.....Jesus Christ and the American Soldier. One died for your soul, the other for your freedom.

duncan
Posts: 10203
Joined: Mon Jul 07, 2008 8:50 pm

Re: 7 Man Syzygy and SSD

Post by duncan » Tue Jul 30, 2019 7:50 pm

syzygy wrote:
Fri Dec 21, 2018 1:27 am

But perfect play once a 6-men ending has been reached on the board is only a small part of the story. The phase before it is much more important (especially at longer time controls where it will indeed be rare that the engine can't figure out a 6-men ending on its own).

My apologies for getting off topic, but is Lomonosov 2 with its 94 Tb Ram capable of solving 8 pieces. ?

https://www.top500.org/system/178444

syzygy
Posts: 4451
Joined: Tue Feb 28, 2012 10:56 pm

Re: 7 Man Syzygy and SSD

Post by syzygy » Tue Jul 30, 2019 10:37 pm

duncan wrote:
Tue Jul 30, 2019 7:50 pm
syzygy wrote:
Fri Dec 21, 2018 1:27 am

But perfect play once a 6-men ending has been reached on the board is only a small part of the story. The phase before it is much more important (especially at longer time controls where it will indeed be rare that the engine can't figure out a 6-men ending on its own).

My apologies for getting off topic, but is Lomonosov 2 with its 94 Tb Ram capable of solving 8 pieces. ?

https://www.top500.org/system/178444
That should be more than enough RAM.

Post Reply