Just reorganised posts 1 and 3 again. The results log is too big now for a single mersenneforum post !

We are now at 40% completion, with a further 23% reserved

[CODE]
Searcher done resd

Thomas11 116 54
danaj 118 83
antonio 65 16
robert44444uk 64 19
pinhodecarlos 28 37
ET_ 5 2
mart_r 3 10
henryzz 1 9

Total 400 230


[/CODE]

completed 4254 - 4264e15, no new records
[CODE]4254-4255e15 110 kgaps, largest 1176 @ 4254560683884506581
4255-4256e15 98 kgaps, largest 1158 @ 4255392859187160013
4256-4257e15 114 kgaps, largest 1216 @ 4256659918861186381
4257-4258e15 117 kgaps, largest 1242 @ 4257232678730035271
4258-4259e15 122 kgaps, largest 1186 @ 4258149732507664927
4259-4260e15 118 kgaps, largest 1200 @ 4259355967821041467
4260-4261e15 115 kgaps, largest 1136 @ 4260173086628254361
4261-4262e15 101 kgaps, largest 1204 @ 4261502609890972783
4262-4263e15 127 kgaps, largest 1232 @ 4262045084255083679
4263-4264e15 104 kgaps, largest 1226 @ 4263188508680031677[/CODE]Reserving 4270-4280e15

[QUOTE=Antonio;461238]I just ran all 3 vesions my i5-3570k (ivybridge) is capable of on my testbed, and the results are identical within the measurement accuracy of the test:
Nehalem:...... 20.27e9 n/sec. 150 wall sec. 554.15 cpu sec. Thread Efficiency: 92.36%
Sandybridge: 20.27e9 n/sec. 149 wall sec. 553.51 cpu sec. Thread Efficiency: 92.29%
Ivybridge:...... 20.27e9 n/sec. 150 wall sec. 554.96 cpu sec. Thread Efficiency: 92.49%
Notes:
Test parameters: -n1 4248786e12 -n2 4248789e12 -gap 1250 -delta 150 -sb 24 -bs 16 -t 4 -mem 12.5
Thread Efficiency = (cpu sec.) / (threads * (wall sec.))

So it would seem that not only is AVX of no significant use, the compiler directives mtune and march have minimal/no effect either. It's the last two that are the most disappointing.[/QUOTE]
I just ran all 4 vesions on my i7-6700k laptop using the same parameters as before, with the following results:
4 Threads:
Nehalem:...... 15.71e9n/sec. 195 wall sec. 702.94 cpu sec. Thread Efficiency: 90.12%
Sandybridge: 15.71e9n/sec. 195 wall sec. 702.91 cpu sec. Thread Efficiency: 90.12%
Ivybridge:...... 15.54e9n/sec. 194 wall sec. 706.28 cpu sec. Thread Efficiency: 91.02%
Haswell..:...... 15.79e9n/sec. 194 wall sec. 703.22 cpu sec. Thread Efficiency: 90.62%

8 Threads:
Nehalem:...... 19.74e9n/sec. 155 wall sec. 1081.19 cpu sec. Thread Efficiency: 87.19%
Sandybridge: 19.74e9n/sec. 156 wall sec. 1087.20 cpu sec. Thread Efficiency: 87.12%
Ivybridge:...... 19.61e9n/sec. 157 wall sec. 1087.81 cpu sec. Thread Efficiency: 86.61%
Haswell..:...... 20.00e9n/sec. 150 wall sec. 1075.44 cpu sec. Thread Efficiency: 87.29%

Strangely the Ivybridge build does badly compared to Nehalem and Sandybridge builds, while there is a definite advantage to using the Haswell build.
In terms of throughput per core (3.9e9/sec for 4 cores, 2.5e9/sec for 8 cores) it looks like it's a waste of resources to use machines with lots of cores and, because the code is trying to make full use of the L3 cache, running multiple instances of the gap code (or other software) on the same machine would have a major impact on performance.

Some statistics of the first 500 e15 ranges

[B]Range 4000 -4100 [/B]
Completeness 94%.
Highest gap achieved 1306 Robert4444UK. Not >1344

[B]Range 4100 -4200[/B]
Completeness 100%
Highest gaps achieved a) 1368 Thomas11 Previously found by LLh in 2011
b) 1380 Thomas11 Previously found by Hertzog and Oliveira Silva in 2007

[B]Range 4200 -4300[/B]
Completeness 64%
Highest gaps achieved a) 1362 Antonio! Previously found by Oliveira Silva in 2011
b) 1372 Antonio! Previously found by LLh in 2014

[B]Range 4300 -4400[/B]
Completeness 100%
Highest gap achieved 1386 dañaj Previously found by Pardi and Oliveira Silva in 2012

[B]Range 4400 -4500[/B]
Completeness 16%
Highest gaps achieved a)1284 Thomas11 Not > 1344
b)1284 pinhodecarlos Not > 1344

From [B]Range 4500 -4999[/B] (Second 500 e15 ranges)
Only gap over 1344 is between 4709 -4710. 1374 by danaj previously found in 2011 by Pardi and Oliveira Silva

[QUOTE=rudy235;461310]Some statistics of the first 500 e15 ranges
[B]Range 4200 -4300[/B]
Completeness 64%
Highest gaps achieved a) 1362 Antonio! Previously found by Oliveira Silva in 2011
b) 1372 Antonio! Previously found by LLh in 2014
[/QUOTE]
You missed out: 1440 @ 4253027105513399527

[QUOTE=Antonio;461323]You missed out: 1440 @ 4253027105513399527[/QUOTE]

I stand corrected.

[B]Range 4200 -4300[/B]
Completeness 64%
Highest gaps achieved a) 1362 Antonio! Previously found by Oliveira Silva in 2011
b) 1372 Antonio! Previously found by LLh in 2014
c) 1440 Antonio! Previously found by LLh in 2014

In theory, with the program output and a name, we could automate the tables and stats. The solutions files would work also, just missing a few details that could be easily provided in a metafile.

I'm thinking of a stats page like the dynamic record prime gaps pages.

[QUOTE=danaj;461327]In theory, with the program output and a name, we could automate the tables and stats. The solutions files would work also, just missing a few details that could be easily provided in a metafile.

I'm thinking of a stats page like the dynamic record prime gaps pages.[/QUOTE]

That's exactly what I had thought. The same way as they do at [url]https://www.mersenne.org/assignments/?exp_lo=71770529&exp_hi=73000000&execm=1&exp1=1&extf=1&exdchk=1[/url] or even at [URL="https://primes.utm.edu/primes/lists/short.pdf"]https://primes.utm.edu/primes/lists/short.pdf[/URL]

Updated #1 and #3

One result from me - my computer had a nervous breakdown yesterday thanks to B Gates incompetence - plus new danaj results.

1238 4077712228247259179

I'm cutting by half my cores running my range since heat wave is coming to UK....nice, always cold...

[QUOTE=Antonio;461288]I just ran all 4 vesions on my i7-6700k laptop using the same parameters as before, with the following results:
4 Threads:
Nehalem:...... 15.71e9n/sec. 195 wall sec. 702.94 cpu sec. Thread Efficiency: 90.12%
Sandybridge: 15.71e9n/sec. 195 wall sec. 702.91 cpu sec. Thread Efficiency: 90.12%
Ivybridge:...... 15.54e9n/sec. 194 wall sec. 706.28 cpu sec. Thread Efficiency: 91.02%
Haswell..:...... 15.79e9n/sec. 194 wall sec. 703.22 cpu sec. Thread Efficiency: 90.62%

8 Threads:
Nehalem:...... 19.74e9n/sec. 155 wall sec. 1081.19 cpu sec. Thread Efficiency: 87.19%
Sandybridge: 19.74e9n/sec. 156 wall sec. 1087.20 cpu sec. Thread Efficiency: 87.12%
Ivybridge:...... 19.61e9n/sec. 157 wall sec. 1087.81 cpu sec. Thread Efficiency: 86.61%
Haswell..:...... 20.00e9n/sec. 150 wall sec. 1075.44 cpu sec. Thread Efficiency: 87.29%

Strangely the Ivybridge build does badly compared to Nehalem and Sandybridge builds, while there is a definite advantage to using the Haswell build.
In terms of throughput per core (3.9e9/sec for 4 cores, 2.5e9/sec for 8 cores) it looks like it's a waste of resources to use machines with lots of cores and, because the code is trying to make full use of the L3 cache, running multiple instances of the gap code (or other software) on the same machine would have a major impact on performance.[/QUOTE]

A 6700K in a laptop? Surely that must be limited in some way. 95 watts is about the peak of my cooling on a desktop. Your timings aren't too different to mine so maybe not.