mersenneforum.org GIMPS Credit for Exponents Between 79.3 Million and 596 Million
 Register FAQ Search Today's Posts Mark Forums Read

2006-11-13, 07:25   #23
jinydu

Dec 2003
Hopefully Near M48

110110111102 Posts

Quote:
 Originally Posted by cheesehead My data? Are you referring to the entries in the "[x]jmptable[p]" tables in Prime95 source module mult.asm?
Yes. Unfortunately, I don't know where the Prime95 source module mult.asm is.

Quote:
 Originally Posted by cheesehead To which of the tables in http://www.mersenneforum.org/showthread.php?t=2731 are you referring?
The first one; post #8 by Joe_O.

So here is an interesting question. Right now, I have 367.584 CPU years of credit for CPU testing ( That figure hasn't budged since Oct. 27 ). How large would an exponent have to single-handedly be worth that much GIMPS credit?

Last fiddled with by jinydu on 2006-11-13 at 07:25

 2006-11-13, 15:19 #24 S485122     "Jacob" Sep 2006 Brussels, Belgium 11·173 Posts On the first page of the GIMPS site you will see a link in the upper right corner : Source Code On that page there is another link : download all the source code(3.6MB)
2006-11-26, 23:08   #25
jinydu

Dec 2003
Hopefully Near M48

2·3·293 Posts

Quote:
 Originally Posted by cheesehead Look in Prime95 source module mult.asm at tables xjmptable and xjmptablep. Some raw excerpts: xjmptable DD 743, 32, 0.00000111, 896 ... allfft DD 68130000, 3670016, 0.323, 9456000 ... allfft DD 77910000, 4194304, 0.382, 5724160 ... allfft DD 96830000, 5242880, 0.485, 7045376 ... allfft DD 115300000, 6291456, 0.668, 8364416 ... allfft DD 134200000, 7340032, 0.886, 9683456 ... xjmptablep DD 739, 32, 0.00000111, 1152 ... allfft DD 58410000, 3145728, 0.289, 4143104 ... allfft DD 77700000, 4194304, 0.425, 5462016 ... DD 115000000, 6291456, 0.668, 9650176 ... DD 153100000, 8388608, 1.042, 12828672 ... I'll leave those to your interpretation. :-)
There are two tables under xjmptable. In each table, the first numerical column appears to be the maximum size of the exponent for that exponent to have that FFT size; the second numerical column is the FFT size. For some reason, the second table seems to be wrong; for instance, it seems to claim a FFT jump at p = 20M, when in fact no such jump exists.

For the first table (which seems to be correct), there seems to be a lot of near (but not exact) repetition for any given FFT range. As the numbers get larger, the table becomes increasingly repetitive. However, the indented text towards right does change a little with each repetition; I'm not sure what that means. There is a third number to the right of the FFT size that stays the same with each repetition of an FFT size; I don't know what it means either. For FFT size 6,291,456, that number is 0.668. It might have something to do with credit; but even if it does, whether or not Primenet awards credit in accordance with this chart is still an open question for me.

So if my guess is correct, the complete list of FFT ranges for version 24.14 are:

Code:
$5\leq p\leq742$                       32
$743\leq p\leq1098$                    48
$1099\leq p\leq1468$                   64
$1469\leq p\leq1826$                   80
$1827\leq p\leq2178$                   96
$2179\leq p\leq2538$                    112
$2539\leq p\leq2904$                   128
$2905\leq p\leq3612$                    160
$3613\leq p\leq4310$                   192
$4311\leq p\leq5028$                   224
$5029\leq p\leq5754$                   256
$5755\leq p\leq7148$                   320
$7149\leq p\leq8526$                   384
$8527\leq p\leq9932$                   448
$9933\leq p\leq11358$                   512
$11359\leq p\leq14118$                  640
$14119\leq p\leq16838$                  768
$16839\leq p\leq19638$                  896
$19639\leq p\leq22476$                  1024
$22477\leq p\leq27898$                 1280
$27899\leq p\leq33288$                  1536
$33289\leq p\leq38798$                  1792
$38799\leq p\leq44338$                 2048
$44339\leq p\leq55098$                 2560
$55099\leq p\leq65728$                  3072
$65729\leq p\leq76558$                  3584
$76559\leq p\leq87548$                 4096
$87549\leq p\leq108799$                5120
$108800\leq p\leq129899$                6144
$129900\leq p\leq151299$                7168
$151300\leq p\leq172699$                8192
$172700\leq p\leq214399$                10240
$214400\leq p\leq255299$                12288
$255300\leq p\leq297299$                14336
$297300\leq p\leq340399$                16384
$340400\leq p\leq423299$               20480
$423300\leq p\leq504599$                24576
$504600\leq p\leq587499$                28672
$587500\leq p\leq671399$                32768
$671400\leq p\leq835199$                40960
$835200\leq p\leq995499$                49152
$995500\leq p\leq1157999$               57344
$1158000\leq p\leq1324999$              65536
$1325000\leq p\leq1647999$              81920
$1648000\leq p\leq1965999$              98304
$1966000\leq p\leq2286999$              114688
$2287000\leq p\leq2613999$              131072
$2614000\leq p\leq3250999$              163840
$3251000\leq p\leq3874999$              196608
$3875000\leq p\leq4511999$              229376
$4512000\leq p\leq5157999$              262144
$5158000\leq p\leq6420999$              327680
$6421000\leq p\leq7650999$              393216
$7651000\leq p\leq8907999$              458752
$8908000\leq p\leq10179999$             524288
$10180000\leq p\leq12649999$            655360
$12650000\leq p\leq15069999$             786432
$15070000\leq p\leq17549999$            917504
Argh! After going through all that work, it appears that the above table is incorrect...

Last fiddled with by jinydu on 2006-11-27 at 00:07

 2006-12-22, 10:54 #26 jinydu     Dec 2003 Hopefully Near M48 33368 Posts StarQwest received 79.417 CPU years for testing M100,000,007. Does anyone know how to account for that figure?

 Similar Threads Thread Thread Starter Forum Replies Last Post E_tron Lounge 55 2007-11-27 15:31 jasong Puzzles 15 2007-08-25 03:01 DJones Lone Mersenne Hunters 5 2007-03-31 19:06 wackyeh Lone Mersenne Hunters 7 2003-07-16 18:42 Prime95 PSearch 4 2003-02-22 08:42

All times are UTC. The time now is 09:06.

Mon Jan 30 09:06:47 UTC 2023 up 165 days, 6:35, 0 users, load averages: 1.00, 0.96, 0.94