PrimeGrid Challenges 2023

[StefanR5R]

At times when there is no competition going on, PrimeGrid's stats of recent average CPU time might be influenced by perhaps a larger percentage of older computers contributing (or vice versa: during competitions, more people bring newer gear online, which will bring average CPU time down), and computers not being tuned very well for the suproject (e.g., many of the active computers are set up to run a mix of different subprojects with settings which are generally more or less suitable, but not quite optimal for the subproject we are looking at).

 

[Markfw]

At times when there is no competition going on, PrimeGrid's stats of recent average CPU time might be influenced by perhaps a larger percentage of older computers contributing (or vice versa: during competitions, more people bring newer gear online, which will bring average CPU time down), and computers not being tuned very well for the suproject (e.g., many of the active computers are set up to run a mix of different subprojects with settings which are generally more or less suitable, but not quite optimal for the subproject we are looking at).

I was only giving my recent times. No idea what the real average is.

 

[StefanR5R]

By "recent average", I meant PrimeGrid's recent average across all hosts of all participants, not particularly your average. :-)

PrimeGrid's average across all hosts should be a lot higher than the times which you are getting, because your computer is fast and seems to be configured at or near the optimum. (Particularly, the optimum for least CPU time per task = for highest host throughput.)

Also, my impression is that many of the computers which are returning 321llr results currently are Intel computers. These have much less cache per core than AMD Zen2…Zen4. Therefore, Intel owners tend to configure fewer concurrent tasks per computer, and correspondingly higher application thread count per task, than AMD owners would choose. Higher thread count -> more internal application overhead -> more CPU time per task.

 

[Markfw]

By "recent average", I meant PrimeGrid's recent average across all hosts of all participants, not particularly your average. :-)

PrimeGrid's average across all hosts should be a lot higher than the times which you are getting, because your computer is fast and seems to be configured at or near the optimum. (Particularly, the optimum for least CPU time per task = for highest host throughput.)

Also, my impression is that many of the computers which are returning 321llr results currently are Intel computers. These have much less cache per core than AMD Zen2…Zen4. Therefore, Intel owners tend to configure fewer concurrent tasks per computer, and correspondingly higher application thread count per task, than AMD owners would choose. Higher thread count -> more internal application overhead -> more CPU time per task.

For my tests, I have been working with @emoga to optimize the 9554. The 7950x's and my 2 Genoa hosts ALL have avx-512, which the PG site specifically says will be the fastest.

But I will be MOST appreciative and looking forward to all you are going to work on. Especially the cpu pinning for linux hosts. I only have lasso, and that only works in windows.

 

[StefanR5R]

@Markfw, when you are in Linux, could you obtain the output of lscpu -e of a Ryzen 7000 and of an EPYC Genoa, and post them here?

(AFAIK Linux uses the same principle for Zen 2 and Zen 3, Ryzen as well as EPYC, to map hardware threads to logical CPUs. But I haven't seen the numbering scheme for Zen 4 yet.)

 

[Markfw]

First a 7950x
mark@mark-X670E-Taichi-linux:~$ lscpu -e
CPU NODE SOCKET CORE L1d:L1i:L2:L3 ONLINE MAXMHZ MINMHZ
0 0 0 0 0:0:0:0 yes 4500.0000 3000.0000
1 0 0 1 1:1:1:0 yes 4500.0000 3000.0000
2 0 0 2 2:2:2:0 yes 4500.0000 3000.0000
3 0 0 3 3:3:3:0 yes 4500.0000 3000.0000
4 0 0 4 4:4:4:0 yes 4500.0000 3000.0000
5 0 0 5 5:5:5:0 yes 4500.0000 3000.0000
6 0 0 6 6:6:6:0 yes 4500.0000 3000.0000
7 0 0 7 7:7:7:0 yes 4500.0000 3000.0000
8 0 0 8 8:8:8:1 yes 4500.0000 3000.0000
9 0 0 9 9:9:9:1 yes 4500.0000 3000.0000
10 0 0 10 10:10:10:1 yes 4500.0000 3000.0000
11 0 0 11 11:11:11:1 yes 4500.0000 3000.0000
12 0 0 12 12:12:12:1 yes 4500.0000 3000.0000
13 0 0 13 13:13:13:1 yes 4500.0000 3000.0000
14 0 0 14 14:14:14:1 yes 4500.0000 3000.0000
15 0 0 15 15:15:15:1 yes 4500.0000 3000.0000
16 0 0 0 0:0:0:0 yes 4500.0000 3000.0000
17 0 0 1 1:1:1:0 yes 4500.0000 3000.0000
18 0 0 2 2:2:2:0 yes 4500.0000 3000.0000
19 0 0 3 3:3:3:0 yes 4500.0000 3000.0000
20 0 0 4 4:4:4:0 yes 4500.0000 3000.0000
21 0 0 5 5:5:5:0 yes 4500.0000 3000.0000
22 0 0 6 6:6:6:0 yes 4500.0000 3000.0000
23 0 0 7 7:7:7:0 yes 4500.0000 3000.0000
24 0 0 8 8:8:8:1 yes 4500.0000 3000.0000
25 0 0 9 9:9:9:1 yes 4500.0000 3000.0000
26 0 0 10 10:10:10:1 yes 4500.0000 3000.0000
27 0 0 11 11:11:11:1 yes 4500.0000 3000.0000
28 0 0 12 12:12:12:1 yes 4500.0000 3000.0000
29 0 0 13 13:13:13:1 yes 4500.0000 3000.0000
30 0 0 14 14:14:14:1 yes 4500.0000 3000.0000
31 0 0 15 15:15:15:1 yes 4500.0000 3000.0000

Now a 96 core Genoa. This mac number goes up and down depending on the application.
mark@EPYC-9654:~$ lscpu -e
CPU NODE SOCKET CORE L1d:L1i:L2:L3 ONLINE MAXMHZ MINMHZ
0 0 0 0 0:0:0:0 yes 2400.0000 1500.0000
1 0 0 1 1:1:1:0 yes 2400.0000 1500.0000
2 0 0 2 2:2:2:0 yes 2400.0000 1500.0000
3 0 0 3 3:3:3:0 yes 2400.0000 1500.0000
4 0 0 4 4:4:4:0 yes 2400.0000 1500.0000
5 0 0 5 5:5:5:0 yes 2400.0000 1500.0000
6 0 0 6 6:6:6:0 yes 2400.0000 1500.0000
7 0 0 7 7:7:7:0 yes 2400.0000 1500.0000
8 0 0 8 8:8:8:1 yes 2400.0000 1500.0000
9 0 0 9 9:9:9:1 yes 2400.0000 1500.0000
10 0 0 10 10:10:10:1 yes 2400.0000 1500.0000
11 0 0 11 11:11:11:1 yes 2400.0000 1500.0000
12 0 0 12 12:12:12:1 yes 2400.0000 1500.0000
13 0 0 13 13:13:13:1 yes 2400.0000 1500.0000
14 0 0 14 14:14:14:1 yes 2400.0000 1500.0000
15 0 0 15 15:15:15:1 yes 2400.0000 1500.0000
16 0 0 16 16:16:16:2 yes 2400.0000 1500.0000
17 0 0 17 17:17:17:2 yes 2400.0000 1500.0000
18 0 0 18 18:18:18:2 yes 2400.0000 1500.0000
19 0 0 19 19:19:19:2 yes 2400.0000 1500.0000
20 0 0 20 20:20:20:2 yes 2400.0000 1500.0000
21 0 0 21 21:21:21:2 yes 2400.0000 1500.0000
22 0 0 22 22:22:22:2 yes 2400.0000 1500.0000
23 0 0 23 23:23:23:2 yes 2400.0000 1500.0000
24 0 0 24 24:24:24:3 yes 2400.0000 1500.0000
25 0 0 25 25:25:25:3 yes 2400.0000 1500.0000
26 0 0 26 26:26:26:3 yes 2400.0000 1500.0000
27 0 0 27 27:27:27:3 yes 2400.0000 1500.0000
28 0 0 28 28:28:28:3 yes 2400.0000 1500.0000
29 0 0 29 29:29:29:3 yes 2400.0000 1500.0000
30 0 0 30 30:30:30:3 yes 2400.0000 1500.0000
31 0 0 31 31:31:31:3 yes 2400.0000 1500.0000
32 0 0 32 32:32:32:4 yes 2400.0000 1500.0000
33 0 0 33 33:33:33:4 yes 2400.0000 1500.0000
34 0 0 34 34:34:34:4 yes 2400.0000 1500.0000
35 0 0 35 35:35:35:4 yes 2400.0000 1500.0000
36 0 0 36 36:36:36:4 yes 2400.0000 1500.0000
37 0 0 37 37:37:37:4 yes 2400.0000 1500.0000
38 0 0 38 38:38:38:4 yes 2400.0000 1500.0000
39 0 0 39 39:39:39:4 yes 2400.0000 1500.0000
40 0 0 40 40:40:40:5 yes 2400.0000 1500.0000
41 0 0 41 41:41:41:5 yes 2400.0000 1500.0000
42 0 0 42 42:42:42:5 yes 2400.0000 1500.0000
43 0 0 43 43:43:43:5 yes 2400.0000 1500.0000
44 0 0 44 44:44:44:5 yes 2400.0000 1500.0000
45 0 0 45 45:45:45:5 yes 2400.0000 1500.0000
46 0 0 46 46:46:46:5 yes 2400.0000 1500.0000
47 0 0 47 47:47:47:5 yes 2400.0000 1500.0000
48 0 0 48 48:48:48:6 yes 2400.0000 1500.0000
49 0 0 49 49:49:49:6 yes 2400.0000 1500.0000
50 0 0 50 50:50:50:6 yes 2400.0000 1500.0000
51 0 0 51 51:51:51:6 yes 2400.0000 1500.0000
52 0 0 52 52:52:52:6 yes 2400.0000 1500.0000
53 0 0 53 53:53:53:6 yes 2400.0000 1500.0000
54 0 0 54 54:54:54:6 yes 2400.0000 1500.0000
55 0 0 55 55:55:55:6 yes 2400.0000 1500.0000
56 0 0 56 56:56:56:7 yes 2400.0000 1500.0000
57 0 0 57 57:57:57:7 yes 2400.0000 1500.0000
58 0 0 58 58:58:58:7 yes 2400.0000 1500.0000
59 0 0 59 59:59:59:7 yes 2400.0000 1500.0000
60 0 0 60 60:60:60:7 yes 2400.0000 1500.0000
61 0 0 61 61:61:61:7 yes 2400.0000 1500.0000
62 0 0 62 62:62:62:7 yes 2400.0000 1500.0000
63 0 0 63 63:63:63:7 yes 2400.0000 1500.0000
64 0 0 64 64:64:64:8 yes 2400.0000 1500.0000
65 0 0 65 65:65:65:8 yes 2400.0000 1500.0000
66 0 0 66 66:66:66:8 yes 2400.0000 1500.0000
67 0 0 67 67:67:67:8 yes 2400.0000 1500.0000
68 0 0 68 68:68:68:8 yes 2400.0000 1500.0000
69 0 0 69 69:69:69:8 yes 2400.0000 1500.0000
70 0 0 70 70:70:70:8 yes 2400.0000 1500.0000
71 0 0 71 71:71:71:8 yes 2400.0000 1500.0000
72 0 0 72 72:72:72:9 yes 2400.0000 1500.0000
73 0 0 73 73:73:73:9 yes 2400.0000 1500.0000
74 0 0 74 74:74:74:9 yes 2400.0000 1500.0000
75 0 0 75 75:75:75:9 yes 2400.0000 1500.0000
76 0 0 76 76:76:76:9 yes 2400.0000 1500.0000
77 0 0 77 77:77:77:9 yes 2400.0000 1500.0000
78 0 0 78 78:78:78:9 yes 2400.0000 1500.0000
79 0 0 79 79:79:79:9 yes 2400.0000 1500.0000
80 0 0 80 80:80:80:10 yes 2400.0000 1500.0000
81 0 0 81 81:81:81:10 yes 2400.0000 1500.0000
82 0 0 82 82:82:82:10 yes 2400.0000 1500.0000
83 0 0 83 83:83:83:10 yes 2400.0000 1500.0000
84 0 0 84 84:84:84:10 yes 2400.0000 1500.0000
85 0 0 85 85:85:85:10 yes 2400.0000 1500.0000
86 0 0 86 86:86:86:10 yes 2400.0000 1500.0000
87 0 0 87 87:87:87:10 yes 2400.0000 1500.0000
88 0 0 88 88:88:88:11 yes 2400.0000 1500.0000
89 0 0 89 89:89:89:11 yes 2400.0000 1500.0000
90 0 0 90 90:90:90:11 yes 2400.0000 1500.0000
91 0 0 91 91:91:91:11 yes 2400.0000 1500.0000
92 0 0 92 92:92:92:11 yes 2400.0000 1500.0000
93 0 0 93 93:93:93:11 yes 2400.0000 1500.0000
94 0 0 94 94:94:94:11 yes 2400.0000 1500.0000
95 0 0 95 95:95:95:11 yes 2400.0000 1500.0000
96 0 0 0 0:0:0:0 yes 2400.0000 1500.0000
97 0 0 1 1:1:1:0 yes 2400.0000 1500.0000
98 0 0 2 2:2:2:0 yes 2400.0000 1500.0000
99 0 0 3 3:3:3:0 yes 2400.0000 1500.0000
100 0 0 4 4:4:4:0 yes 2400.0000 1500.0000
101 0 0 5 5:5:5:0 yes 2400.0000 1500.0000
102 0 0 6 6:6:6:0 yes 2400.0000 1500.0000
103 0 0 7 7:7:7:0 yes 2400.0000 1500.0000
104 0 0 8 8:8:8:1 yes 2400.0000 1500.0000
105 0 0 9 9:9:9:1 yes 2400.0000 1500.0000
106 0 0 10 10:10:10:1 yes 2400.0000 1500.0000
107 0 0 11 11:11:11:1 yes 2400.0000 1500.0000
108 0 0 12 12:12:12:1 yes 2400.0000 1500.0000
109 0 0 13 13:13:13:1 yes 2400.0000 1500.0000
110 0 0 14 14:14:14:1 yes 2400.0000 1500.0000
111 0 0 15 15:15:15:1 yes 2400.0000 1500.0000
112 0 0 16 16:16:16:2 yes 2400.0000 1500.0000
113 0 0 17 17:17:17:2 yes 2400.0000 1500.0000
114 0 0 18 18:18:18:2 yes 2400.0000 1500.0000
115 0 0 19 19:19:19:2 yes 2400.0000 1500.0000
116 0 0 20 20:20:20:2 yes 2400.0000 1500.0000
117 0 0 21 21:21:21:2 yes 2400.0000 1500.0000
118 0 0 22 22:22:22:2 yes 2400.0000 1500.0000
119 0 0 23 23:23:23:2 yes 2400.0000 1500.0000
120 0 0 24 24:24:24:3 yes 2400.0000 1500.0000
121 0 0 25 25:25:25:3 yes 2400.0000 1500.0000
122 0 0 26 26:26:26:3 yes 2400.0000 1500.0000
123 0 0 27 27:27:27:3 yes 2400.0000 1500.0000
124 0 0 28 28:28:28:3 yes 2400.0000 1500.0000
125 0 0 29 29:29:29:3 yes 2400.0000 1500.0000
126 0 0 30 30:30:30:3 yes 2400.0000 1500.0000
127 0 0 31 31:31:31:3 yes 2400.0000 1500.0000
128 0 0 32 32:32:32:4 yes 2400.0000 1500.0000
129 0 0 33 33:33:33:4 yes 2400.0000 1500.0000
130 0 0 34 34:34:34:4 yes 2400.0000 1500.0000
131 0 0 35 35:35:35:4 yes 2400.0000 1500.0000
132 0 0 36 36:36:36:4 yes 2400.0000 1500.0000
133 0 0 37 37:37:37:4 yes 2400.0000 1500.0000
134 0 0 38 38:38:38:4 yes 2400.0000 1500.0000
135 0 0 39 39:39:39:4 yes 2400.0000 1500.0000
136 0 0 40 40:40:40:5 yes 2400.0000 1500.0000
137 0 0 41 41:41:41:5 yes 2400.0000 1500.0000
138 0 0 42 42:42:42:5 yes 2400.0000 1500.0000
139 0 0 43 43:43:43:5 yes 2400.0000 1500.0000
140 0 0 44 44:44:44:5 yes 2400.0000 1500.0000
141 0 0 45 45:45:45:5 yes 2400.0000 1500.0000
142 0 0 46 46:46:46:5 yes 2400.0000 1500.0000
143 0 0 47 47:47:47:5 yes 2400.0000 1500.0000
144 0 0 48 48:48:48:6 yes 2400.0000 1500.0000
145 0 0 49 49:49:49:6 yes 2400.0000 1500.0000
146 0 0 50 50:50:50:6 yes 2400.0000 1500.0000
147 0 0 51 51:51:51:6 yes 2400.0000 1500.0000
148 0 0 52 52:52:52:6 yes 2400.0000 1500.0000
149 0 0 53 53:53:53:6 yes 2400.0000 1500.0000
150 0 0 54 54:54:54:6 yes 2400.0000 1500.0000
151 0 0 55 55:55:55:6 yes 2400.0000 1500.0000
152 0 0 56 56:56:56:7 yes 2400.0000 1500.0000
153 0 0 57 57:57:57:7 yes 2400.0000 1500.0000
154 0 0 58 58:58:58:7 yes 2400.0000 1500.0000
155 0 0 59 59:59:59:7 yes 2400.0000 1500.0000
156 0 0 60 60:60:60:7 yes 2400.0000 1500.0000
157 0 0 61 61:61:61:7 yes 2400.0000 1500.0000
158 0 0 62 62:62:62:7 yes 2400.0000 1500.0000
159 0 0 63 63:63:63:7 yes 2400.0000 1500.0000
160 0 0 64 64:64:64:8 yes 2400.0000 1500.0000
161 0 0 65 65:65:65:8 yes 2400.0000 1500.0000
162 0 0 66 66:66:66:8 yes 2400.0000 1500.0000
163 0 0 67 67:67:67:8 yes 2400.0000 1500.0000
164 0 0 68 68:68:68:8 yes 2400.0000 1500.0000
165 0 0 69 69:69:69:8 yes 2400.0000 1500.0000
166 0 0 70 70:70:70:8 yes 2400.0000 1500.0000
167 0 0 71 71:71:71:8 yes 2400.0000 1500.0000
168 0 0 72 72:72:72:9 yes 2400.0000 1500.0000
169 0 0 73 73:73:73:9 yes 2400.0000 1500.0000
170 0 0 74 74:74:74:9 yes 2400.0000 1500.0000
171 0 0 75 75:75:75:9 yes 2400.0000 1500.0000
172 0 0 76 76:76:76:9 yes 2400.0000 1500.0000
173 0 0 77 77:77:77:9 yes 2400.0000 1500.0000
174 0 0 78 78:78:78:9 yes 2400.0000 1500.0000
175 0 0 79 79:79:79:9 yes 2400.0000 1500.0000
176 0 0 80 80:80:80:10 yes 2400.0000 1500.0000
177 0 0 81 81:81:81:10 yes 2400.0000 1500.0000
178 0 0 82 82:82:82:10 yes 2400.0000 1500.0000
179 0 0 83 83:83:83:10 yes 2400.0000 1500.0000
180 0 0 84 84:84:84:10 yes 2400.0000 1500.0000
181 0 0 85 85:85:85:10 yes 2400.0000 1500.0000
182 0 0 86 86:86:86:10 yes 2400.0000 1500.0000
183 0 0 87 87:87:87:10 yes 2400.0000 1500.0000
184 0 0 88 88:88:88:11 yes 2400.0000 1500.0000
185 0 0 89 89:89:89:11 yes 2400.0000 1500.0000
186 0 0 90 90:90:90:11 yes 2400.0000 1500.0000
187 0 0 91 91:91:91:11 yes 2400.0000 1500.0000
188 0 0 92 92:92:92:11 yes 2400.0000 1500.0000
189 0 0 93 93:93:93:11 yes 2400.0000 1500.0000
190 0 0 94 94:94:94:11 yes 2400.0000 1500.0000
191 0 0 95 95:95:95:11 yes 2400.0000 1500.0000

 

[cellarnoise]

First a 7950x
mark@mark-X670E-Taichi-linux:~$ lscpu -e
CPU NODE SOCKET CORE L1d:L1i:L2:L3 ONLINE MAXMHZ MINMHZ
0 0 0 0 0:0:0:0 yes 4500.0000 3000.0000
1 0 0 1 1:1:1:0 yes 4500.0000 3000.0000
2 0 0 2 2:2:2:0 yes 4500.0000 3000.0000
3 0 0 3 3:3:3:0 yes 4500.0000 3000.0000
4 0 0 4 4:4:4:0 yes 4500.0000 3000.0000
5 0 0 5 5:5:5:0 yes 4500.0000 3000.0000
6 0 0 6 6:6:6:0 yes 4500.0000 3000.0000
7 0 0 7 7:7:7:0 yes 4500.0000 3000.0000
8 0 0 8 8:8:8:1 yes 4500.0000 3000.0000
9 0 0 9 9:9:9:1 yes 4500.0000 3000.0000
10 0 0 10 10:10:10:1 yes 4500.0000 3000.0000
11 0 0 11 11:11:11:1 yes 4500.0000 3000.0000
12 0 0 12 12:12:12:1 yes 4500.0000 3000.0000
13 0 0 13 13:13:13:1 yes 4500.0000 3000.0000
14 0 0 14 14:14:14:1 yes 4500.0000 3000.0000
15 0 0 15 15:15:15:1 yes 4500.0000 3000.0000
16 0 0 0 0:0:0:0 yes 4500.0000 3000.0000
17 0 0 1 1:1:1:0 yes 4500.0000 3000.0000
18 0 0 2 2:2:2:0 yes 4500.0000 3000.0000
19 0 0 3 3:3:3:0 yes 4500.0000 3000.0000
20 0 0 4 4:4:4:0 yes 4500.0000 3000.0000
21 0 0 5 5:5:5:0 yes 4500.0000 3000.0000
22 0 0 6 6:6:6:0 yes 4500.0000 3000.0000
23 0 0 7 7:7:7:0 yes 4500.0000 3000.0000
24 0 0 8 8:8:8:1 yes 4500.0000 3000.0000
25 0 0 9 9:9:9:1 yes 4500.0000 3000.0000
26 0 0 10 10:10:10:1 yes 4500.0000 3000.0000
27 0 0 11 11:11:11:1 yes 4500.0000 3000.0000
28 0 0 12 12:12:12:1 yes 4500.0000 3000.0000
29 0 0 13 13:13:13:1 yes 4500.0000 3000.0000
30 0 0 14 14:14:14:1 yes 4500.0000 3000.0000
31 0 0 15 15:15:15:1 yes 4500.0000 3000.0000

Now a 96 core Genoa. This mac number goes up and down depending on the application.
mark@EPYC-9654:~$ lscpu -e
CPU NODE SOCKET CORE L1d:L1i:L2:L3 ONLINE MAXMHZ MINMHZ
0 0 0 0 0:0:0:0 yes 2400.0000 1500.0000
1 0 0 1 1:1:1:0 yes 2400.0000 1500.0000
2 0 0 2 2:2:2:0 yes 2400.0000 1500.0000
3 0 0 3 3:3:3:0 yes 2400.0000 1500.0000
4 0 0 4 4:4:4:0 yes 2400.0000 1500.0000
5 0 0 5 5:5:5:0 yes 2400.0000 1500.0000
6 0 0 6 6:6:6:0 yes 2400.0000 1500.0000
7 0 0 7 7:7:7:0 yes 2400.0000 1500.0000
8 0 0 8 8:8:8:1 yes 2400.0000 1500.0000
9 0 0 9 9:9:9:1 yes 2400.0000 1500.0000
10 0 0 10 10:10:10:1 yes 2400.0000 1500.0000
11 0 0 11 11:11:11:1 yes 2400.0000 1500.0000
12 0 0 12 12:12:12:1 yes 2400.0000 1500.0000
13 0 0 13 13:13:13:1 yes 2400.0000 1500.0000
14 0 0 14 14:14:14:1 yes 2400.0000 1500.0000
15 0 0 15 15:15:15:1 yes 2400.0000 1500.0000
16 0 0 16 16:16:16:2 yes 2400.0000 1500.0000
17 0 0 17 17:17:17:2 yes 2400.0000 1500.0000
18 0 0 18 18:18:18:2 yes 2400.0000 1500.0000
19 0 0 19 19:19:19:2 yes 2400.0000 1500.0000
20 0 0 20 20:20:20:2 yes 2400.0000 1500.0000
21 0 0 21 21:21:21:2 yes 2400.0000 1500.0000
22 0 0 22 22:22:22:2 yes 2400.0000 1500.0000
23 0 0 23 23:23:23:2 yes 2400.0000 1500.0000
24 0 0 24 24:24:24:3 yes 2400.0000 1500.0000
25 0 0 25 25:25:25:3 yes 2400.0000 1500.0000
26 0 0 26 26:26:26:3 yes 2400.0000 1500.0000
27 0 0 27 27:27:27:3 yes 2400.0000 1500.0000
28 0 0 28 28:28:28:3 yes 2400.0000 1500.0000
29 0 0 29 29:29:29:3 yes 2400.0000 1500.0000
30 0 0 30 30:30:30:3 yes 2400.0000 1500.0000
31 0 0 31 31:31:31:3 yes 2400.0000 1500.0000
32 0 0 32 32:32:32:4 yes 2400.0000 1500.0000
33 0 0 33 33:33:33:4 yes 2400.0000 1500.0000
34 0 0 34 34:34:34:4 yes 2400.0000 1500.0000
35 0 0 35 35:35:35:4 yes 2400.0000 1500.0000
36 0 0 36 36:36:36:4 yes 2400.0000 1500.0000
37 0 0 37 37:37:37:4 yes 2400.0000 1500.0000
38 0 0 38 38:38:38:4 yes 2400.0000 1500.0000
39 0 0 39 39:39:39:4 yes 2400.0000 1500.0000
40 0 0 40 40:40:40:5 yes 2400.0000 1500.0000
41 0 0 41 41:41:41:5 yes 2400.0000 1500.0000
42 0 0 42 42:42:42:5 yes 2400.0000 1500.0000
43 0 0 43 43:43:43:5 yes 2400.0000 1500.0000
44 0 0 44 44:44:44:5 yes 2400.0000 1500.0000
45 0 0 45 45:45:45:5 yes 2400.0000 1500.0000
46 0 0 46 46:46:46:5 yes 2400.0000 1500.0000
47 0 0 47 47:47:47:5 yes 2400.0000 1500.0000
48 0 0 48 48:48:48:6 yes 2400.0000 1500.0000
49 0 0 49 49:49:49:6 yes 2400.0000 1500.0000
50 0 0 50 50:50:50:6 yes 2400.0000 1500.0000
51 0 0 51 51:51:51:6 yes 2400.0000 1500.0000
52 0 0 52 52:52:52:6 yes 2400.0000 1500.0000
53 0 0 53 53:53:53:6 yes 2400.0000 1500.0000
54 0 0 54 54:54:54:6 yes 2400.0000 1500.0000
55 0 0 55 55:55:55:6 yes 2400.0000 1500.0000
56 0 0 56 56:56:56:7 yes 2400.0000 1500.0000
57 0 0 57 57:57:57:7 yes 2400.0000 1500.0000
58 0 0 58 58:58:58:7 yes 2400.0000 1500.0000
59 0 0 59 59:59:59:7 yes 2400.0000 1500.0000
60 0 0 60 60:60:60:7 yes 2400.0000 1500.0000
61 0 0 61 61:61:61:7 yes 2400.0000 1500.0000
62 0 0 62 62:62:62:7 yes 2400.0000 1500.0000
63 0 0 63 63:63:63:7 yes 2400.0000 1500.0000
64 0 0 64 64:64:64:8 yes 2400.0000 1500.0000
65 0 0 65 65:65:65:8 yes 2400.0000 1500.0000
66 0 0 66 66:66:66:8 yes 2400.0000 1500.0000
67 0 0 67 67:67:67:8 yes 2400.0000 1500.0000
68 0 0 68 68:68:68:8 yes 2400.0000 1500.0000
69 0 0 69 69:69:69:8 yes 2400.0000 1500.0000
70 0 0 70 70:70:70:8 yes 2400.0000 1500.0000
71 0 0 71 71:71:71:8 yes 2400.0000 1500.0000
72 0 0 72 72:72:72:9 yes 2400.0000 1500.0000
73 0 0 73 73:73:73:9 yes 2400.0000 1500.0000
74 0 0 74 74:74:74:9 yes 2400.0000 1500.0000
75 0 0 75 75:75:75:9 yes 2400.0000 1500.0000
76 0 0 76 76:76:76:9 yes 2400.0000 1500.0000
77 0 0 77 77:77:77:9 yes 2400.0000 1500.0000
78 0 0 78 78:78:78:9 yes 2400.0000 1500.0000
79 0 0 79 79:79:79:9 yes 2400.0000 1500.0000
80 0 0 80 80:80:80:10 yes 2400.0000 1500.0000
81 0 0 81 81:81:81:10 yes 2400.0000 1500.0000
82 0 0 82 82:82:82:10 yes 2400.0000 1500.0000
83 0 0 83 83:83:83:10 yes 2400.0000 1500.0000
84 0 0 84 84:84:84:10 yes 2400.0000 1500.0000
85 0 0 85 85:85:85:10 yes 2400.0000 1500.0000
86 0 0 86 86:86:86:10 yes 2400.0000 1500.0000
87 0 0 87 87:87:87:10 yes 2400.0000 1500.0000
88 0 0 88 88:88:88:11 yes 2400.0000 1500.0000
89 0 0 89 89:89:89:11 yes 2400.0000 1500.0000
90 0 0 90 90:90:90:11 yes 2400.0000 1500.0000
91 0 0 91 91:91:91:11 yes 2400.0000 1500.0000
92 0 0 92 92:92:92:11 yes 2400.0000 1500.0000
93 0 0 93 93:93:93:11 yes 2400.0000 1500.0000
94 0 0 94 94:94:94:11 yes 2400.0000 1500.0000
95 0 0 95 95:95:95:11 yes 2400.0000 1500.0000
96 0 0 0 0:0:0:0 yes 2400.0000 1500.0000
97 0 0 1 1:1:1:0 yes 2400.0000 1500.0000
98 0 0 2 2:2:2:0 yes 2400.0000 1500.0000
99 0 0 3 3:3:3:0 yes 2400.0000 1500.0000
100 0 0 4 4:4:4:0 yes 2400.0000 1500.0000
101 0 0 5 5:5:5:0 yes 2400.0000 1500.0000
102 0 0 6 6:6:6:0 yes 2400.0000 1500.0000
103 0 0 7 7:7:7:0 yes 2400.0000 1500.0000
104 0 0 8 8:8:8:1 yes 2400.0000 1500.0000
105 0 0 9 9:9:9:1 yes 2400.0000 1500.0000
106 0 0 10 10:10:10:1 yes 2400.0000 1500.0000
107 0 0 11 11:11:11:1 yes 2400.0000 1500.0000
108 0 0 12 12:12:12:1 yes 2400.0000 1500.0000
109 0 0 13 13:13:13:1 yes 2400.0000 1500.0000
110 0 0 14 14:14:14:1 yes 2400.0000 1500.0000
111 0 0 15 15:15:15:1 yes 2400.0000 1500.0000
112 0 0 16 16:16:16:2 yes 2400.0000 1500.0000
113 0 0 17 17:17:17:2 yes 2400.0000 1500.0000
114 0 0 18 18:18:18:2 yes 2400.0000 1500.0000
115 0 0 19 19:19:19:2 yes 2400.0000 1500.0000
116 0 0 20 20:20:20:2 yes 2400.0000 1500.0000
117 0 0 21 21:21:21:2 yes 2400.0000 1500.0000
118 0 0 22 22:22:22:2 yes 2400.0000 1500.0000
119 0 0 23 23:23:23:2 yes 2400.0000 1500.0000
120 0 0 24 24:24:24:3 yes 2400.0000 1500.0000
121 0 0 25 25:25:25:3 yes 2400.0000 1500.0000
122 0 0 26 26:26:26:3 yes 2400.0000 1500.0000
123 0 0 27 27:27:27:3 yes 2400.0000 1500.0000
124 0 0 28 28:28:28:3 yes 2400.0000 1500.0000
125 0 0 29 29:29:29:3 yes 2400.0000 1500.0000
126 0 0 30 30:30:30:3 yes 2400.0000 1500.0000
127 0 0 31 31:31:31:3 yes 2400.0000 1500.0000
128 0 0 32 32:32:32:4 yes 2400.0000 1500.0000
129 0 0 33 33:33:33:4 yes 2400.0000 1500.0000
130 0 0 34 34:34:34:4 yes 2400.0000 1500.0000
131 0 0 35 35:35:35:4 yes 2400.0000 1500.0000
132 0 0 36 36:36:36:4 yes 2400.0000 1500.0000
133 0 0 37 37:37:37:4 yes 2400.0000 1500.0000
134 0 0 38 38:38:38:4 yes 2400.0000 1500.0000
135 0 0 39 39:39:39:4 yes 2400.0000 1500.0000
136 0 0 40 40:40:40:5 yes 2400.0000 1500.0000
137 0 0 41 41:41:41:5 yes 2400.0000 1500.0000
138 0 0 42 42:42:42:5 yes 2400.0000 1500.0000
139 0 0 43 43:43:43:5 yes 2400.0000 1500.0000
140 0 0 44 44:44:44:5 yes 2400.0000 1500.0000
141 0 0 45 45:45:45:5 yes 2400.0000 1500.0000
142 0 0 46 46:46:46:5 yes 2400.0000 1500.0000
143 0 0 47 47:47:47:5 yes 2400.0000 1500.0000
144 0 0 48 48:48:48:6 yes 2400.0000 1500.0000
145 0 0 49 49:49:49:6 yes 2400.0000 1500.0000
146 0 0 50 50:50:50:6 yes 2400.0000 1500.0000
147 0 0 51 51:51:51:6 yes 2400.0000 1500.0000
148 0 0 52 52:52:52:6 yes 2400.0000 1500.0000
149 0 0 53 53:53:53:6 yes 2400.0000 1500.0000
150 0 0 54 54:54:54:6 yes 2400.0000 1500.0000
151 0 0 55 55:55:55:6 yes 2400.0000 1500.0000
152 0 0 56 56:56:56:7 yes 2400.0000 1500.0000
153 0 0 57 57:57:57:7 yes 2400.0000 1500.0000
154 0 0 58 58:58:58:7 yes 2400.0000 1500.0000
155 0 0 59 59:59:59:7 yes 2400.0000 1500.0000
156 0 0 60 60:60:60:7 yes 2400.0000 1500.0000
157 0 0 61 61:61:61:7 yes 2400.0000 1500.0000
158 0 0 62 62:62:62:7 yes 2400.0000 1500.0000
159 0 0 63 63:63:63:7 yes 2400.0000 1500.0000
160 0 0 64 64:64:64:8 yes 2400.0000 1500.0000
161 0 0 65 65:65:65:8 yes 2400.0000 1500.0000
162 0 0 66 66:66:66:8 yes 2400.0000 1500.0000
163 0 0 67 67:67:67:8 yes 2400.0000 1500.0000
164 0 0 68 68:68:68:8 yes 2400.0000 1500.0000
165 0 0 69 69:69:69:8 yes 2400.0000 1500.0000
166 0 0 70 70:70:70:8 yes 2400.0000 1500.0000
167 0 0 71 71:71:71:8 yes 2400.0000 1500.0000
168 0 0 72 72:72:72:9 yes 2400.0000 1500.0000
169 0 0 73 73:73:73:9 yes 2400.0000 1500.0000
170 0 0 74 74:74:74:9 yes 2400.0000 1500.0000
171 0 0 75 75:75:75:9 yes 2400.0000 1500.0000
172 0 0 76 76:76:76:9 yes 2400.0000 1500.0000
173 0 0 77 77:77:77:9 yes 2400.0000 1500.0000
174 0 0 78 78:78:78:9 yes 2400.0000 1500.0000
175 0 0 79 79:79:79:9 yes 2400.0000 1500.0000
176 0 0 80 80:80:80:10 yes 2400.0000 1500.0000
177 0 0 81 81:81:81:10 yes 2400.0000 1500.0000
178 0 0 82 82:82:82:10 yes 2400.0000 1500.0000
179 0 0 83 83:83:83:10 yes 2400.0000 1500.0000
180 0 0 84 84:84:84:10 yes 2400.0000 1500.0000
181 0 0 85 85:85:85:10 yes 2400.0000 1500.0000
182 0 0 86 86:86:86:10 yes 2400.0000 1500.0000
183 0 0 87 87:87:87:10 yes 2400.0000 1500.0000
184 0 0 88 88:88:88:11 yes 2400.0000 1500.0000
185 0 0 89 89:89:89:11 yes 2400.0000 1500.0000
186 0 0 90 90:90:90:11 yes 2400.0000 1500.0000
187 0 0 91 91:91:91:11 yes 2400.0000 1500.0000
188 0 0 92 92:92:92:11 yes 2400.0000 1500.0000
189 0 0 93 93:93:93:11 yes 2400.0000 1500.0000
190 0 0 94 94:94:94:11 yes 2400.0000 1500.0000
191 0 0 95 95:95:95:11 yes 2400.0000 1500.0000

I love mack and cheese... This looks like a lot! Maybe? I hope it is anyway 😁 for boinc? You are all crazy 🤣

 

[TennesseeTony]

I'm going to sit out this round, I don't have any AVX-512 hardware or equivalent. It would be like running Universe on Windows, or a GPU project on CPU. Just not an efficient use of my hardware this time around. Have fun!!

 

[Markfw]

and now the 9554 in linux

mark@9554-2:~$ lscpu -e
CPU NODE SOCKET CORE L1d:L1i:L2:L3 ONLINE MAXMHZ MINMHZ
0 0 0 0 0:0:0:0 yes 2900.0000 1500.0000
1 0 0 1 1:1:1:0 yes 2900.0000 1500.0000
2 0 0 2 2:2:2:0 yes 2900.0000 1500.0000
3 0 0 3 3:3:3:0 yes 2900.0000 1500.0000
4 0 0 4 4:4:4:0 yes 2900.0000 1500.0000
5 0 0 5 5:5:5:0 yes 2900.0000 1500.0000
6 0 0 6 6:6:6:0 yes 2900.0000 1500.0000
7 0 0 7 7:7:7:0 yes 2900.0000 1500.0000
8 0 0 8 8:8:8:1 yes 2900.0000 1500.0000
9 0 0 9 9:9:9:1 yes 2900.0000 1500.0000
10 0 0 10 10:10:10:1 yes 2900.0000 1500.0000
11 0 0 11 11:11:11:1 yes 2900.0000 1500.0000
12 0 0 12 12:12:12:1 yes 2900.0000 1500.0000
13 0 0 13 13:13:13:1 yes 2900.0000 1500.0000
14 0 0 14 14:14:14:1 yes 2900.0000 1500.0000
15 0 0 15 15:15:15:1 yes 2900.0000 1500.0000
16 0 0 16 16:16:16:2 yes 2900.0000 1500.0000
17 0 0 17 17:17:17:2 yes 2900.0000 1500.0000
18 0 0 18 18:18:18:2 yes 2900.0000 1500.0000
19 0 0 19 19:19:19:2 yes 2900.0000 1500.0000
20 0 0 20 20:20:20:2 yes 2900.0000 1500.0000
21 0 0 21 21:21:21:2 yes 2900.0000 1500.0000
22 0 0 22 22:22:22:2 yes 2900.0000 1500.0000
23 0 0 23 23:23:23:2 yes 2900.0000 1500.0000
24 0 0 24 24:24:24:3 yes 2900.0000 1500.0000
25 0 0 25 25:25:25:3 yes 2900.0000 1500.0000
26 0 0 26 26:26:26:3 yes 2900.0000 1500.0000
27 0 0 27 27:27:27:3 yes 2900.0000 1500.0000
28 0 0 28 28:28:28:3 yes 2900.0000 1500.0000
29 0 0 29 29:29:29:3 yes 2900.0000 1500.0000
30 0 0 30 30:30:30:3 yes 2900.0000 1500.0000
31 0 0 31 31:31:31:3 yes 2900.0000 1500.0000
32 0 0 32 32:32:32:4 yes 2900.0000 1500.0000
33 0 0 33 33:33:33:4 yes 2900.0000 1500.0000
34 0 0 34 34:34:34:4 yes 2900.0000 1500.0000
35 0 0 35 35:35:35:4 yes 2900.0000 1500.0000
36 0 0 36 36:36:36:4 yes 2900.0000 1500.0000
37 0 0 37 37:37:37:4 yes 2900.0000 1500.0000
38 0 0 38 38:38:38:4 yes 2900.0000 1500.0000
39 0 0 39 39:39:39:4 yes 2900.0000 1500.0000
40 0 0 40 40:40:40:5 yes 2900.0000 1500.0000
41 0 0 41 41:41:41:5 yes 2900.0000 1500.0000
42 0 0 42 42:42:42:5 yes 2900.0000 1500.0000
43 0 0 43 43:43:43:5 yes 2900.0000 1500.0000
44 0 0 44 44:44:44:5 yes 2900.0000 1500.0000
45 0 0 45 45:45:45:5 yes 2900.0000 1500.0000
46 0 0 46 46:46:46:5 yes 2900.0000 1500.0000
47 0 0 47 47:47:47:5 yes 2900.0000 1500.0000
48 0 0 48 48:48:48:6 yes 2900.0000 1500.0000
49 0 0 49 49:49:49:6 yes 2900.0000 1500.0000
50 0 0 50 50:50:50:6 yes 2900.0000 1500.0000
51 0 0 51 51:51:51:6 yes 2900.0000 1500.0000
52 0 0 52 52:52:52:6 yes 2900.0000 1500.0000
53 0 0 53 53:53:53:6 yes 2900.0000 1500.0000
54 0 0 54 54:54:54:6 yes 2900.0000 1500.0000
55 0 0 55 55:55:55:6 yes 2900.0000 1500.0000
56 0 0 56 56:56:56:7 yes 2900.0000 1500.0000
57 0 0 57 57:57:57:7 yes 2900.0000 1500.0000
58 0 0 58 58:58:58:7 yes 2900.0000 1500.0000
59 0 0 59 59:59:59:7 yes 2900.0000 1500.0000
60 0 0 60 60:60:60:7 yes 2900.0000 1500.0000
61 0 0 61 61:61:61:7 yes 2900.0000 1500.0000
62 0 0 62 62:62:62:7 yes 2900.0000 1500.0000
63 0 0 63 63:63:63:7 yes 2900.0000 1500.0000
64 0 0 0 0:0:0:0 yes 2900.0000 1500.0000
65 0 0 1 1:1:1:0 yes 2900.0000 1500.0000
66 0 0 2 2:2:2:0 yes 2900.0000 1500.0000
67 0 0 3 3:3:3:0 yes 2900.0000 1500.0000
68 0 0 4 4:4:4:0 yes 2900.0000 1500.0000
69 0 0 5 5:5:5:0 yes 2900.0000 1500.0000
70 0 0 6 6:6:6:0 yes 2900.0000 1500.0000
71 0 0 7 7:7:7:0 yes 2900.0000 1500.0000
72 0 0 8 8:8:8:1 yes 2900.0000 1500.0000
73 0 0 9 9:9:9:1 yes 2900.0000 1500.0000
74 0 0 10 10:10:10:1 yes 2900.0000 1500.0000
75 0 0 11 11:11:11:1 yes 2900.0000 1500.0000
76 0 0 12 12:12:12:1 yes 2900.0000 1500.0000
77 0 0 13 13:13:13:1 yes 2900.0000 1500.0000
78 0 0 14 14:14:14:1 yes 2900.0000 1500.0000
79 0 0 15 15:15:15:1 yes 2900.0000 1500.0000
80 0 0 16 16:16:16:2 yes 2900.0000 1500.0000
81 0 0 17 17:17:17:2 yes 2900.0000 1500.0000
82 0 0 18 18:18:18:2 yes 2900.0000 1500.0000
83 0 0 19 19:19:19:2 yes 2900.0000 1500.0000
84 0 0 20 20:20:20:2 yes 2900.0000 1500.0000
85 0 0 21 21:21:21:2 yes 2900.0000 1500.0000
86 0 0 22 22:22:22:2 yes 2900.0000 1500.0000
87 0 0 23 23:23:23:2 yes 2900.0000 1500.0000
88 0 0 24 24:24:24:3 yes 2900.0000 1500.0000
89 0 0 25 25:25:25:3 yes 2900.0000 1500.0000
90 0 0 26 26:26:26:3 yes 2900.0000 1500.0000
91 0 0 27 27:27:27:3 yes 2900.0000 1500.0000
92 0 0 28 28:28:28:3 yes 2900.0000 1500.0000
93 0 0 29 29:29:29:3 yes 2900.0000 1500.0000
94 0 0 30 30:30:30:3 yes 2900.0000 1500.0000
95 0 0 31 31:31:31:3 yes 2900.0000 1500.0000
96 0 0 32 32:32:32:4 yes 2900.0000 1500.0000
97 0 0 33 33:33:33:4 yes 2900.0000 1500.0000
98 0 0 34 34:34:34:4 yes 2900.0000 1500.0000
99 0 0 35 35:35:35:4 yes 2900.0000 1500.0000
100 0 0 36 36:36:36:4 yes 2900.0000 1500.0000
101 0 0 37 37:37:37:4 yes 2900.0000 1500.0000
102 0 0 38 38:38:38:4 yes 2900.0000 1500.0000
103 0 0 39 39:39:39:4 yes 2900.0000 1500.0000
104 0 0 40 40:40:40:5 yes 2900.0000 1500.0000
105 0 0 41 41:41:41:5 yes 2900.0000 1500.0000
106 0 0 42 42:42:42:5 yes 2900.0000 1500.0000
107 0 0 43 43:43:43:5 yes 2900.0000 1500.0000
108 0 0 44 44:44:44:5 yes 2900.0000 1500.0000
109 0 0 45 45:45:45:5 yes 2900.0000 1500.0000
110 0 0 46 46:46:46:5 yes 2900.0000 1500.0000
111 0 0 47 47:47:47:5 yes 2900.0000 1500.0000
112 0 0 48 48:48:48:6 yes 2900.0000 1500.0000
113 0 0 49 49:49:49:6 yes 2900.0000 1500.0000
114 0 0 50 50:50:50:6 yes 2900.0000 1500.0000
115 0 0 51 51:51:51:6 yes 2900.0000 1500.0000
116 0 0 52 52:52:52:6 yes 2900.0000 1500.0000
117 0 0 53 53:53:53:6 yes 2900.0000 1500.0000
118 0 0 54 54:54:54:6 yes 2900.0000 1500.0000
119 0 0 55 55:55:55:6 yes 2900.0000 1500.0000
120 0 0 56 56:56:56:7 yes 2900.0000 1500.0000
121 0 0 57 57:57:57:7 yes 2900.0000 1500.0000
122 0 0 58 58:58:58:7 yes 2900.0000 1500.0000
123 0 0 59 59:59:59:7 yes 2900.0000 1500.0000
124 0 0 60 60:60:60:7 yes 2900.0000 1500.0000
125 0 0 61 61:61:61:7 yes 2900.0000 1500.0000
126 0 0 62 62:62:62:7 yes 2900.0000 1500.0000
127 0 0 63 63:63:63:7 yes 2900.0000 1500.0000

 

[Markfw]

For Mondays PG event. I now have 224 Zen 4 Genoa cores and 96 7950x Zen 4 cores available ! I am all Zen 4 now !
and I will have lasso on 64 Genoa cores, and 48 7950x cores.

 

[cellarnoise]

I'm going to sit out this round, I don't have any AVX-512 hardware or equivalent. It would be like running Universe on Windows, or a GPU project on CPU. Just not an efficient use of my hardware this time around. Have fun!!

Anything with even avx-256 would do o.k.?

 

[StefanR5R]

Anything with even avx-256 would do o.k.?

I agree.

The Haswell-E 5820K with 15 MB last-level cache could run 1 task at once, 6 software thread/task. It has got FMA3 support, only drawback is its rather old process node (22 nm), which means high Watt/PPD.

Linux would do well in keeping the 6 software threads on all 6 physical cores most of the time, out of the box. I'm not so sure about Windows; it might need some tweaking for this, e.g. disabling HT in the BIOS.

The Zen 2 CPUs on 7 nm and FMA3 support still work very efficiently for PrimeGrid's LLR2 based projects, especially for ones like llr321 which fit into the cache of one Zen 2 core complex. Ryzen 9 3900X with two 6-core CCXs, 16 MB last-level cache per CCX, would most likely work best with 2 tasks running at once, 6 software threads/task. Edit, oops, wrong, see post #189.

It is important for LLR2's efficiency on Zen 2…4 though that the software threads of each task are not spread across CCX boundaries. There are a few Windows-specific solutions for this out there, as well as at least one Linux-specific solution, plus one OS-independent solution involving multiple BOINC client instances, plus one OS-independent solution which only works on EPYCs. I have been using the latter two types myself so far, but hope to get another Linux-specifc solution pieced together which shall be more convenient to handle.

One can of course use Zen 2…4 without the CCX related optimization and it would still do rather well. But still, the performance and power efficiency uplift with this optimization is quite sizable.

 

[Ken g6]

Let's count down to the challenge...3...2....1...oh, it's already started!

 

[StefanR5R]

The Zen 2 CPUs on 7 nm and FMA3 support still work very efficiently for PrimeGrid's LLR2 based projects, especially for ones like llr321 which fit into the cache of one Zen 2 core complex. Ryzen 9 3900X with two 6-core CCXs, 16 MB last-level cache per CCX, would most likely work best with 2 tasks running at once, 6 software threads/task.

Argh, I wasn't thinking straight. Ryzen 9 3900X has got four 3-core CCXs of course. It's true that there are 16 MB last-level cache per each 3c/6t CCX. It's likely that four 3-threaded llr321 tasks, if pinned to the proper logical CPUs, would be able to fully utilize the FMA units (one AVX-256 FMA unit per physical core), and thus already provide the throughput optimum. Task duration should be less than 8 hours 9 hours in this case. (Extrapolating from my Epyc with 4 cores per CCX but slower clock.)

 

[Ken g6]

Day 1 stats:

Rank___Credits____Username
4______2478163____xii5ku
5______2187128____markfw
7______2163315____crashtech
9______1693024____[TA]Skillz
10_____1666917____w a h
18_____790424_____biodoc
24_____525508_____cellarnoise2
52_____230759_____Orange Kid
76_____132367_____waffleironhead
98_____89518______Pokey
110____87974______Ken_g6
146____52991______Letin Noxe
168____33426______mmonnin

Rank__Credits____Team
1_____12131521___TeAm AnandTech
2_____8524544____Antarctic Crunchers
3_____7244510____Czech National Team
4_____5169027____BOINC@AUSTRALIA

I'd had cool weather where I live until this started, on the hottest day of the year so far.

 

[StefanR5R]

If you log into your account at the PrimeGrid web site and go to "Edit preferences" via "PrimeGrid preferences", you'll see [...] that current 321-LLR work takes on average 44 hours of CPU time (averaged over all recent returns from all kinds of computers; applies to "main tasks" only of course, not the tiny verification tasks),

By "recent average", I meant PrimeGrid's recent average across all hosts of all participants, not particularly your average. :-)

PrimeGrid's average across all hosts should be a lot higher than the times which you are getting, because your computer is fast and seems to be configured at or near the optimum. (Particularly, the optimum for least CPU time per task = for highest host throughput.)

Also, my impression is that many of the computers which are returning 321llr results currently are Intel computers. These have much less cache per core than AMD Zen2…Zen4. Therefore, Intel owners tend to configure fewer concurrent tasks per computer, and correspondingly higher application thread count per task, than AMD owners would choose. Higher thread count -> more internal application overhead -> more CPU time per task.

Another influence on PrimeGrid's recent average CPU time per task is this: The fraction of results-returning computers which are configured to use HyperThreading/ SMT for LLR2 tasks. (Using HT/SMT sometimes brings a small throughput advantage, sometimes a small throughput disadvantage.¹) If one and the same computer ran one and the same workunit with and without HT/SMT, it would spend circa double the CPU time for a task with HT/SMT, compared to without HT/SMT, as far as I have seen. — Of course it is impossible for us to figure out what the fractions of HT/SMT vs. non-HT/SMT configured computers within PrimeGrid's recent average CPU time per task are.

I have two computer types going myself:
The way I configured the first one, it spends ≈73 h CPU time within ≈7.2 h elapsed time per task.
The other type spends merely ≈34 h CPU time within ≈8.7 h elapsed time per task.
Type one is using HT, type two is not.

________
¹) I spoke of a possible small throughput disadvantage. Of course it could become a bigger throughput disadvantage as soon as there is not enough CPU cache for the overall workload on a computer.

 

[Ken g6]

Day 2 stats:

Rank___Credits____Username
5______5686214____xii5ku
6______5228369____[TA]Skillz
7______4936495____markfw
8______4753435____crashtech
9______3929094____w a h
17_____2112475____cellarnoise2
20_____1788711____biodoc
53_____541003_____Orange Kid
73_____343521_____waffleironhead
113____210732_____Ken_g6
124____178437_____Pokey
129____176145_____mmonnin
183____87795______Letin Noxe
247____22087______woochowski
269____11080______DRFUNGUS

Rank__Credits____Team
1_____30005599___TeAm AnandTech
2_____21037701___Antarctic Crunchers
3_____14493232___Czech National Team
4_____12158542___BOINC@AUSTRALIA

Welcome to some lesser-known names joining us late. I don't even recognize DRFUNGUS.

 

[Skillz]

Drfungus is one of the gamers that play on my rust server. He's uses really, really old hardware. Being a teenager that's all he can do for now.

He's on the teams discord server.

 

[Ken g6]

Day 4 stats:

Rank___Credits____Username
5______12636211___[TA]Skillz
6______11811581___xii5ku
7______10740967___markfw
8______9442266____crashtech
9______8009801____w a h
17_____4352119____cellarnoise2
20_____3699544____biodoc
54_____1210857____Orange Kid
81_____645274_____waffleironhead
119____446500_____Pokey
126____401793_____Ken_g6
161____265149_____mmonnin
205____135271_____Letin Noxe
255____54974______woochowski
265____44414______DRFUNGUS

Rank__Credits____Team
1_____63896730___TeAm AnandTech
2_____48767517___Antarctic Crunchers
3_____29223987___BOINC@AUSTRALIA
4_____28856356___Czech National Team

Yesterday I ran stats but forgot to post them!

 

[Markfw]

I am writing from Trout creek Montana. Glad I am holding 7th

 

[Ken g6]

More-or-less final stats:

Rank___Credits____Username
5______16480880___[TA]Skillz
6______15039421___xii5ku
7______13754312___markfw
8______11636166___crashtech
10_____9861975____w a h
18_____5550188____cellarnoise2
22_____4600440____biodoc
55_____1560104____Orange Kid
90_____790522_____waffleironhead
102____663727_____Pokey
125____526045_____Ken_g6
178____271133_____Letin Noxe
182____265149_____mmonnin
262____66488______DRFUNGUS
264____66100______woochowski

Rank__Credits____Team
1_____81132657___TeAm AnandTech
2_____66486850___Antarctic Crunchers
3_____42357002___Czech National Team
4_____40282410___BOINC@AUSTRALIA

Alas, Venec sneaked into 9th place over @emoga, so we only have an arithmetic progression of four ranks.

 

[StefanR5R]

Thanks for the stats! :-)

The TeAm's progress into the 2023 season so far:

Jan	Mar	Apr	Jun	Jul	Aug	Sep	Oct	Dec
CPU GPU​	CPU​	CPU​	CPU​	GPU​	GPU​	CPU​	CPU​	GPU CPU​
1st​	1st​	1st​	1st​	tbd​	tbd​	tbd​	tbd​	tbd​

 

[Orange Kid]

Thanks for all the stats. The usual middle of the local pack for me. 🙂

 

[cellarnoise]

Thanks for all the stats. The usual middle of the local pack for me. 🙂

Mid-packers bring up the middle of this team!

 

[Markfw]

@Ken g6 , what kind of video card does best at the upcoming challenge ? Like a 3090 or a Titan V type ?