I'm not a programmer at all, so don't know if Robert G's software can be adjusted to do E28 to E37. Of course, you can always use Axn's software, which is known to be reliable,and which we used before Robert G's work.

The only issue there is the fact that Axn's software is not as efficient, maybe by a factor of 5 to 10 at finding VPS. But is perfectly adequate for smaller k targets. It is probably not the most efficient algorithm for searching for long CCs.

I do have the E28 and E37 Reisel racing records chart - no work was carried out on the Sierpinski side.

Regards

Robert

I have already considered some additional modifications to the program (e.g. adding the check whether a sequence is actually a member of a higher E level, automatic and optimal choice of Nash and Smith check parameters, etc.). And, perhaps, adding a graphical user interface, which would make it more attractive to potential new coworkers...

Right now, I'm not sure, if the algorithm can be easily changed for handling also E=28 and E=36. There are some hard-coded multipliers, which probably must changed.

Nevertheless, it's good to know that there is increasing interest in this program. This gives me some additional push forward...

[QUOTE=Thomas11;292704]There was also a 116/10000 the same day:
[CODE]S 144614850311421 60 100/5496 116/10000 K=1206288612570998158177981005 iteration=50 I=22614 Sun Mar 11 08:48:44 2012[/CODE]

I will start sieving them (together with some other 110+ sequences) this evening. Stay tuned...[/QUOTE]

Oh well, it turns out that this sequence is actually S 215814850311421 66, which I found already earlier.

But there is another noteworthy sequence in the news:

S 387050125955097 172 100/8526 105/10000

Actually the first S172!

With the new enhance of having the possibility to stop the client at a given iteration/subiteration it is more easy to split the work into different clients. This means various co-workers can work with the same E therefore I think it is time to create a reservation table.

[QUOTE=robert44444uk;292628]Before we receive more details of Thomas11's latest candidate... [/QUOTE]

Up to n=20,000 it produced 128 primes, which is the second best sequences at this level (after 134/20000 for the S66 found earlier)

And it set a few new absolute records, including a noteworthy [B]115/7000[/B]:
[CODE]111 5954 216650161657053 S 58
112 6392 216650161657053 S 58
113 6395 216650161657053 S 58
114 6527 216650161657053 S 58
115 6907 216650161657053 S 58
116 8489 216650161657053 S 58

118 8998 216650161657053 S 58
119 9097 216650161657053 S 58
120 9805 216650161657053 S 58

122 11389 216650161657053 S 58
[/CODE]

Gosh page 17 of this thread suddenly.

Some more absolute Riesel records:

159 139973 440310850049907 R 82
160 143571 440310850049907 R 82

Three more for fresh territory on the Riesel side.

Congrats Thomas11 for all his new records!

[QUOTE=Thomas11;292884]Up to n=20,000 it produced 128 primes, which is the second best sequences at this level (after 134/20000 for the S66 found earlier)

And it set a few new absolute records, including a noteworthy [B]115/7000[/B]:
[CODE]111 5954 216650161657053 S 58
112 6392 216650161657053 S 58
113 6395 216650161657053 S 58
114 6527 216650161657053 S 58
115 6907 216650161657053 S 58
116 8489 216650161657053 S 58

118 8998 216650161657053 S 58
119 9097 216650161657053 S 58
120 9805 216650161657053 S 58

122 11389 216650161657053 S 58
[/CODE][/QUOTE]

Don't think they are all absolute...but they are certainly Sierpinski records. My 121 Riesel candidate produced:

112 6159
113 6213
114 6218
115 6872
116 7388
117 7523
118 7600
119 7799
120 9283

But Mr Riesel concedes 111 and 122, where the best are 5996 and 12865 (from a E106 candidate)

Mr Riesel also claims one E66 at 128/19034

Encouraged by Gary's and Carlos' request I had a closer look into the source code. It turns out that it can be modified for E<52 quite easily.

Attached is a version for E>=36. There is a 32bit (for P4) and a 64bit binary (for Core2).

I did a few tests for the new code and it seems to work quite well.
And it already produced a nice Sierpinski VPS (maybe already known):

S 177693671 36 100/7398 106/10000 K=7770564667602165

Please note the following facts:
For E=36 I had to reduce the size of the CRT cycle. The cycle of the original code is based on the primes 7, 17, 23, 41, 47, 71, 79, and 97, whose product yields a cycle length of 2869549272527.
For E=36 we need to skip 79 and 97. This reduces the length of each cycle (iteration) to 374468129. This means that the iterations are much shorter now, and that there are only 3024 sub-iterations (I's) compared to 120960 for the original code.

The code should also work for higher E levels (52, 58, and the like).
But it will be significantly slower than the original code.

I will post another version (for E>=28) in a separate post.

Please do some tests and report about problems you may encounter.
Depending on your cpu power it would be nice if someone could verify at least one E>=52 VPS with the E=36 code...

And here comes the E>=28 variant.
The cycle length is again shorter, since we need to skip also p=71 (Ord2=35).
This results in a cycle length of 5274199 with only 84 sub-iterations (I) per iteration.

Again, this version will be slower than the E>=36 and the E>=52 variants.

Of course, in a future version we should combine all those variants (also for E=10/12) in one single executable. But we need to find a proper solution for distinguishing between the different iteration/I schemes (e.g. when running "one" iteration for E=52 which different cycle lengths).

I forgot to include the source code. So here it is...

And another milestone is reached:
There are now more than [B]10,000 known Sierpinski VPS[/B]! :smile:

Here is a little statistics:
[CODE] Total M52 M58 M60 M66 M82 M100 M106 M130 M138 M148 M162 M172 M178 M180 M196 M210 M226
100 2683 580 659 316 637 118 167 105 69 17 13 0 0 2 0 0 0 0
101 2155 442 531 274 529 88 129 89 48 7 17 0 0 0 0 0 1 0
102 1770 339 436 211 502 62 101 68 32 6 13 0 0 0 0 0 0 0
103 1274 278 295 169 311 54 74 48 29 6 9 1 0 0 0 0 0 0
104 886 153 216 108 259 34 41 38 31 3 3 0 0 0 0 0 0 0
105 673 132 135 79 209 32 40 26 13 3 3 0 1 0 0 0 0 0
106 401 68 95 53 127 15 21 11 6 2 2 0 0 1 0 0 0 0
107 287 48 63 35 89 11 23 14 1 1 2 0 0 0 0 0 0 0
108 180 39 35 21 48 9 13 10 1 2 2 0 0 0 0 0 0 0
109 106 12 26 11 34 7 8 4 2 0 2 0 0 0 0 0 0 0
110 64 10 16 10 16 3 6 3 0 0 0 0 0 0 0 0 0 0
111 40 6 9 5 12 4 4 0 0 0 0 0 0 0 0 0 0 0
112 21 1 4 1 9 0 1 2 1 1 1 0 0 0 0 0 0 0
113 13 3 2 0 5 0 1 2 0 0 0 0 0 0 0 0 0 0
114 9 3 0 1 5 0 0 0 0 0 0 0 0 0 0 0 0 0
115 4 1 1 0 2 0 0 0 0 0 0 0 0 0 0 0 0 0
116 1 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0
117 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
118 1 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0
119 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
120 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
total: 10570 2116 2524 1294 2796 437 629 420 233 48 67 1 1 3 0 0 1 0

iterations: 58 70 32 80 52 2000 2290 3100 3100 6850 2435 600 285 660 1350 800 800
VPS/iter: 36.5 36.1 40.4 35.0 8.4 0.3 0.2 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
[/CODE]

Actually I have no explanation for the slightly higher outcome for E=60. Maybe this is due to different levels in the Smith check.