| Author |
Message |
|---|
 RogerVolunteer developer
Volunteer tester
Send message
Joined: 27 Nov 11
Posts: 1138
ID: 120786
Credit: 268,621,444
RAC: 0
                    
|
|
27 & 121 combined - Foundations of Computational Mathematics Challenge - 9-16 December
Announced is a challenge at the 27 & 121 prime search projects from the 9th till 16th of December. I suggest we start at 12:00 UTC and end at 12:00 UTC seven days later.
As with previous combined challenges the rankings will be determined by Score. Best to figure out which project your rigs are most efficient at before the challenge starts.
The Society for the Foundations of Computational Mathematics hold a Conference every 3 years in different corners of the globe. This time the Conference is being held in Montevideo, from 11th till 20th of December. What better way to celebrate the Conference than by finding a 27121 prime!
More about the 27121 prime search can be found here. News and infos about the PRPNet client can be found here.
To take part, you have to activate the following lines in prpclient.ini:
server=27:50:1:prpnet.primegrid.com:12006
server=121:50:1:prpnet.primegrid.com:12001
Note the 50 here sets the % of WU sent from the server, set to what %'s you prefer but must sum to 100.
Stats will be done manually. I will take a screenshot at the start and end of the challenge and post everyone's final stats.
All previous PRPNet challenge stats can be found here.
Good luck!
____________
 |
|
|
RogerVolunteer developer
Volunteer tester
Send message
Joined: 27 Nov 11
Posts: 1138
ID: 120786
Credit: 268,621,444
RAC: 0
|
|
Start your Engines!
Green flag waved and the Challenge has started.
I've manually saved the challenge stats on my Hard Drive. Copy in the thread:
http://www.primegrid.com/forum_thread.php?id=6022 |
|
|
Michael Goetz Volunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 3,564
                          
|
server=27:50:1:prpnet.primegrid.com:12006
server=121:50:1:prpnet.primegrid.com:12001
Note the 50 here sets the % of WU sent from the server, set to what %'s you prefer but must sum to 100.
A long time ago, in a galaxy far, far away -- or at least a long time ago -- I was writing a stock trading system and was working on the entry screen that allowed users to allocate percentages of stocks. It worked in a similar fashion to this PRPNet configuration; the user just had to enter the percentage he wanted to go into each account.
The problem with that is when you want to evenly divide the stock equally between three accounts. You can't, in base ten, express 1/3 exactly. 33% is wrong. 33.33% is wrong. 33.333333333333333333333333% is wrong. No matter how many 3s you type, the number is still wrong. Not wrong by much, but in the securities industry, they care a LOT of getting the numbers exactly right. The pennies tend to add up really, really quickly.
The solution was simple: the sum of the percentages don't have to add up to 100. Instead of percentages, they're just relative weights. If all are "33", then each gets an equal share. If all are "1234", all get an equal share. If all are "1", all get an equal share.
You may recognize this as the same system that BOINC uses to let you allocate processing time between different projects. You can assign any numbers you want and it doesn't matter what they add up to. If Project A has a weight of 200 and project B has a weight of 100, then project A will get twice as much processing time as project B.
The PRPNet "percentage" numbers were changed a while ago to be arbitrary weights rather than percentages. You can still think of them as percentages and make them add up to 100 if you want to, but they do not need to total 100 any more. Unless you're running a fairly old version of the PRPNet client, you can make the numbers whatever you wish.
____________
My lucky number is 75898524288+1 |
|
|
|
|
|
Michael Goetz, yes, I also recognize that situation very well.
However, even after you allow the numbers (weights) to sum up to anything, so the user can specify any commensurable ratio, you still run into issues when you try to split the amount (or stock count, etc.) according to the weights. For example splitting $100.00 in the ratio 2:1 involves rounding. When splitting into multiple groups, the rounding leads to problems similar to the mathematics of voting systems trying to achieve proportional representation.
/JeppeSN
|
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 3,564
|
Michael Goetz, yes, I also recognize that situation very well.
However, even after you allow the numbers (weights) to sum up to anything, so the user can specify any commensurable ratio, you still run into issues when you try to split the amount (or stock count, etc.) according to the weights. For example splitting $100.00 in the ratio 2:1 involves rounding. When splitting into multiple groups, the rounding leads to problems similar to the mathematics of voting systems trying to achieve proportional representation.
/JeppeSN
That's very true, of course, and the algorithm that does the actual split has to deal with that. The purpose of this method of entry is to make it easier for people to enter their intention without having to pull out a calculator.
____________
My lucky number is 75898524288+1 |
|
|
Scott BrownVolunteer moderator
Project administrator
Volunteer tester
Project scientist
Send message
Joined: 17 Oct 05
Posts: 2329
ID: 1178
Credit: 15,614,402,663
RAC: 11,545,211
                                          
|
...For example splitting $100.00 in the ratio 2:1 involves rounding. When splitting into multiple groups, the rounding leads to problems similar to the mathematics of voting systems trying to achieve proportional representation.
/JeppeSN
Actually, this isn't quite accurate. Both the monetary example and the voting example above describe single discrete events. Thus, they would apply to only a single selection of applications to compute on PRPnet (or more accurately, a single cross-section of what is being computed at a single time point). Crunching on PRPnet is usually done over time, which should appeal to the asymptotic property of such a proportional weighting. That is, at any given time point, the actual mix of applications being computed would typically not match the intended ratio, but over time the overall mix would match the ratio the longer one engaged in computation.
|
|
|
|
|
Actually, this isn't quite accurate. Both the monetary example and the voting example above describe single discrete events. [...]
I agree. PRPNet could approach any ratio (between subprojects), given enough time, as closely as one desires. /JeppeSN |
|
|
|
|
Stats will be done manually. I will take a screenshot at the start and end of the challenge and post everyone's final stats.
I miss the good old stats at http://u-g-f.de/PRPNet/challenge_stats.php. Why are stats manual this time. Is Sysadm@Nbg too busy to manage the stats?
/JeppeSN |
|
|
|
|
Stats will be done manually. I will take a screenshot at the start and end of the challenge and post everyone's final stats.
I miss the good old stats at http://u-g-f.de/PRPNet/challenge_stats.php. Why are stats manual this time. Is Sysadm@Nbg too busy to manage the stats?
Yes: http://www.primegrid.com/forum_thread.php?id=5261&nowrap=true#80877 |
|
|
|
|
Stats will be done manually. I will take a screenshot at the start and end of the challenge and post everyone's final stats.
I miss the good old stats at http://u-g-f.de/PRPNet/challenge_stats.php. Why are stats manual this time. Is Sysadm@Nbg too busy to manage the stats?
Yes: http://www.primegrid.com/forum_thread.php?id=5261&nowrap=true#80877
Anyone else able to get the running stats for the challenge?
____________
|
|
|
JimBHonorary cruncher
Send message
Joined: 4 Aug 11
Posts: 918
ID: 107307
Credit: 977,945,376
RAC: 34
      
|
|
The PRPNet database doesn't record when work is returned. Without having taken a snapshot of the data at the start of the challenge, nobody's going to be able to generate accurate stats. BOINC is different, every task has a sent time and a received time. |
|
|
RogerVolunteer developer
Volunteer tester
Send message
Joined: 27 Nov 11
Posts: 1138
ID: 120786
Credit: 268,621,444
RAC: 0
|
|
Project 27 12/12/14
Users 23:38:04 UTC Challenge Since Start
Total Achieved Start
Lumiukko 53955464 48737560 5217904
brinktastee 33498482 28706641 4791841
zlodeck 4358823 0 4358823
Nortech 56957297 53546042 3411255
composite 169250911 166793715 2457196
Cyph3r 5933248 4433077 1500171
Scott_Brown 169229666 167791368 1438298
Grebuloner 1130003 0 1130003
vanklein 26597217 25774992 822225
GDB 760915 0 760915
Robert_Hoffman 3784883 3209524 575359
Tapio_Rajala 13390056 12896986 493070
JeppeSN 390907 0 390907
Roger 4734044 4365541 368503
SysadmAtNbg 7967706 7659407 308299
TroubledBunny 226232 0 226232
ReRada 323460 118088 205372
serhjo 164728 0 164728
[AF>France]_Bernard_Dagorn 123349 0 123349
101575 82218 0 82218
BernhardFrey 61561 0 61561
Boogyman_Munster 20546 0 20546
Eridanus 20492 0 20492 |
|
|
RogerVolunteer developer
Volunteer tester
Send message
Joined: 27 Nov 11
Posts: 1138
ID: 120786
Credit: 268,621,444
RAC: 0
|
|
Project 27 12/12/14
Teams 23:38:16 UTC Challenge Since Start
Total Achieved Start
PrimeSearchTeam 242460935 237243030 5217905
Aggie_The_Pew 189027980 184236139 4791841
Russia 66626849 62268025 4358824
Portugal@Home 4046756 2546584 1500172
Duke_University 169229666 167791368 1438298
The_Knights_Who_Say_Ni! 1888598 758595 1130003
12121search 9501146 8678921 822225
SETI.Germany 316335801 315822131 513670
Overclockers_Australia 4734044 4365541 368503
UK_BOINC_Team 226232 0 226232
Ukraine 30181728 30017000 164728
L'Alliance_Francophone 2306910 2183561 123349
US_Navy 20546 0 20546
314 20492 0 20492 |
|
|
RogerVolunteer developer
Volunteer tester
Send message
Joined: 27 Nov 11
Posts: 1138
ID: 120786
Credit: 268,621,444
RAC: 0
|
|
Project 121 13/12/14
Users 00:06:04 UTC Challenge Since Start
Total Achieved Start
composite 130552993 88162532 42390461
Lumiukko 107651841 102608688 5043153
Roger 6914670 5360570 1554100
zlodeck 1228997 0 1228997
13678 136650565 135557607 1092958
Grebuloner 1090687 0 1090687
GDB 682258 0 682258
LarManDude 77082012 76400482 681530
Cyph3r 4026438 3447193 579245
Nasicus 111062647 110623475 439172
Buckey 409113 0 409113
SysadmAtNbg 12215642 11908481 307161
serhjo 273204 0 273204
JeppeSN 239000 0 239000
rover 44832097 44593567 238530
TroubledBunny 204836 0 204836
brinktastee 10623530 10419221 204309
Boogyman_Munster 203886 0 203886
[AF>France]_Bernard_Dagorn 170340 0 170340
juhehe 2946175 2810473 135702
ReRada 728324 660157 68167
Eridanus 68008 0 68008 |
|
|
RogerVolunteer developer
Volunteer tester
Send message
Joined: 27 Nov 11
Posts: 1138
ID: 120786
Credit: 268,621,444
RAC: 0
|
|
Project 121 13/12/14
Teams 00:06:11 UTC Challenge Since Start
Total Achieved Start
PrimeSearchTeam 287292953 282249800 5043153
Overclockers_Australia 6914670 5360570 1554100
Russia 60296797 59067801 1228996
The_Knights_Who_Say_Ni! 73658693 72568006 1090687
Portugal@Home 2017441 1438197 579244
Crunchers@Freiburg 112972489 112533317 439172
BOINC_Confederation 426284 17170 409114
SETI.Germany 46899909 46524581 375328
Ukraine 1598721 1325517 273204
Team_Germany 44832097 44593567 238530
UK_BOINC_Team 204836 0 204836
Aggie_The_Pew 387688685 387484377 204308
US_Navy 203886 0 203886
L'Alliance_Francophone 3125495 2955155 170340
314 68008 0 68008 |
|
|
RogerVolunteer developer
Volunteer tester
Send message
Joined: 27 Nov 11
Posts: 1138
ID: 120786
Credit: 268,621,444
RAC: 0
|
|
Current User Combined
Rank Total Achieved
1 composite 44847657
2 Lumiukko 10261057
3 zlodeck 5587820
4 brinktastee 4996150
5 Nortech 3411255
6 Grebuloner 2220690
7 Cyph3r 2079416
8 Roger 1922603
9 GDB 1443173
10 Scott_Brown 1438298
11 13678 1092958
12 vanklein 822225
13 LarManDude 681530
14 JeppeSN 629907
15 SysadmAtNbg 615460
16 Robert_Hoffman 575359
17 Tapio_Rajala 493070
18 Nasicus 439172
19 serhjo 437932
20 TroubledBunny 431068
21 Buckey 409113
22 [AF>France]_Bernard_Dagorn 293689
23 ReRada 273539
24 rover 238530
25 Boogyman_Munster 224432
26 juhehe 135702
27 Eridanus 88500
28 101575 82218
29 BernhardFrey 61561
Current Team Combined
Rank Total Achieved
1 PrimeSearchTeam 10261058
2 Russia 5587820
3 Aggie_The_Pew 4996149
4 The_Knights_Who_Say_Ni! 2220690
5 Portugal@Home 2079416
6 Overclockers_Australia 1922603
7 Duke_University 1438298
8 SETI.Germany 888998
9 12121search 822225
10 Crunchers@Freiburg 439172
11 Ukraine 437932
12 UK_BOINC_Team 431068
13 BOINC_Confederation 409114
14 L'Alliance_Francophone 293689
15 Team_Germany 238530
16 US_Navy 224432
17 314 88500 |
|
|
RogerVolunteer developer
Volunteer tester
Send message
Joined: 27 Nov 11
Posts: 1138
ID: 120786
Credit: 268,621,444
RAC: 0
|
|
The current leading edge for the 27 prime search now up to n=4,776,130 on both the plus and minus sides.
The leading edge was at n=4,754,362 at the beginning of the challenge.
Since the challenge started, the leading edge has advanced 0.45% as much as it had prior to the challenge!
The current leading edge for the 121 prime search now up to n=6,152,961 on both the plus and minus sides.
The leading edge was at n=6,107,197 at the beginning of the challenge.
Since the challenge started, the leading edge has advanced 0.74% as much as it had prior to the challenge! |
|
|
RogerVolunteer developer
Volunteer tester
Send message
Joined: 27 Nov 11
Posts: 1138
ID: 120786
Credit: 268,621,444
RAC: 0
|
|
Checkered Flag is waved. Challenge is finished!
Final scores are being counted. |
|
|
RogerVolunteer developer
Volunteer tester
Send message
Joined: 27 Nov 11
Posts: 1138
ID: 120786
Credit: 268,621,444
RAC: 0
|
|
Stats are final. Congratulation to composite and PrimeSearchTeam.
The top of the combined challenge rankings is as follows:
rank userid score
1 composite 81,652,406
2 Lumiukko 21,887,082
3 zlodeck 11,771,433
4 Nortech 6,440,593
5 brinktastee 5,514,645
rank teamid score
1 PrimeSearchTeam 22,079,674
2 Russia 11,771,433
3 Aggie_The_Pew 7,838,146
No prime was found during the challenge however the 27 leading edge went from n=4,754,362 to 4,801,307, that's an increase of 0.98% over where it was at the start of the challenge.
121 leading edge went from n=6,107,197 to 6,160,676, that's an increase of 0.87% over where it was at the start of the challenge.
Congratulations to everyone who participated 2014!
Make a suggestion for the name of a challenge in 2015:
http://www.primegrid.com/forum_thread.php?id=5941&nowrap=true#80452 |
|
|
|
|
|
Will there be complete individual and team standings posted somewhere? (I realize we're in a weird situation with sysadm@nbg being away...) Thanks for keeping things running!
G |
|
|
RogerVolunteer developer
Volunteer tester
Send message
Joined: 27 Nov 11
Posts: 1138
ID: 120786
Credit: 268,621,444
RAC: 0
|
|
Final User Combined
Rank Total Achieved
1 composite 81652406
2 Lumiukko 21887082
3 zlodeck 11771433
4 Nortech 6440593
5 brinktastee 5514645
6 Grebuloner 5058516
7 Cyph3r 4498947
8 GDB 3422644
9 Roger 2997384
10 13678 2734812
11 Scott_Brown 2720378
12 Gary_Craig 2323502
13 vanklein 1693616
14 JeppeSN 1504295
15 serhjo 1442625
16 LarManDude 1365365
17 SysadmAtNbg 1254627
18 Robert_Hoffman 1135303
19 Tapio_Rajala 989867
20 Nasicus 952119
21 TroubledBunny 864009
22 [AF>France]_Bernard_Dagorn 842879
23 ReRada 816201
24 Boogyman_Munster 532566
25 rover 443937
26 Buckey 409113
27 Robert47 192591
28 Eridanus 163987
29 BernhardFrey 144167
30 juhehe 135702
31 101575 82218
Final Team Combined
Rank Total Achieved
1 PrimeSearchTeam 22079674
2 Russia 11771433
3 Aggie_The_Pew 7838146
4 The_Knights_Who_Say_Ni! 5058516
5 Portugal@Home 4498948
6 Overclockers_Australia 2997384
7 Duke_University 2720378
8 SETI.Germany 2070828
9 12121search 1693616
10 Ukraine 1442625
11 UK_BOINC_Team 1143439
12 Crunchers@Freiburg 952119
13 L'Alliance_Francophone 842879
14 US_Navy 532566
15 Team_Germany 443937
16 BOINC_Confederation 409114
17 314 163987 |
|
|
