FMA3 LLR Testing Thread

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

FMA3 LLR Test Plan

(You can help if you want.)

Before we put LLR into production we are going to do extensive testing. Since 6 different builds are involved (32/64 bit and Mac/Linux/Windows), all tests will need to be done for all 6 builds. In addition, with the 64 bit builds the testing will be done on SSE3, AVX (Sandy Bridge or Ivy Bridge), and FMA3 (Haswell) CPUs. In all, 216 tests are required for full validation.

When all of the tests are completed for a particular build it will be put into production. Some builds will therefore go into production before others. While this may seem unfair, each of you may, at any time, use the new app with app_info. In fact, we encourage this as it will help with testing.

If you run the new LLR (3.8.13) on any of the test cases listed below, let me know about it and include the workunit ID so I can track and record it as it validates.

The following projects need to be tested:

PPSE
SoB
SR5 (with c = +1)
SGS
WOO
SR5 (with c = -1)

Each of those 6 needs to be tested on Windows, Linux, and Mac, and each of those 18 combinations needs to be tested on a 32 bit CPU, a 64 bit SSE3 CPU, a 64 bit AVX (SB or IB) CPU, and a 64 bit FMA3 CPU (Haswell).

That's a total of 72 tests, and each test has three parts. The first part is running a task on BOINC that gets accepted by the server, and the second part is when the task gets validated by the server against the normal production application (i.e., not another app_info/anonymous platform app).

The third part is manually running the new LLR against a known prime to prove that it can find primes correctly and produces the correct output. If we're lucky, someone will actually find a PPSE or SGS prime with the new LLR during the testing.

The Windows SSE3, FMA3, and 32 bit testing I can do myself. I'll need someone else to do the Windows AVX tests.

If need be, I should be able to do the Linux SSE3, FMA3, and 32 bit testing in a VM, but if someone else can do it on a native Linux machine that would be preferable. I can't do the Linux AVX tests.

I will, of course, need someone else to do all of the Mac testing. I'm not sure if we're going to find a 32 bit Mac, so we may want to test the 32 bit Mac app on a 64 bit Mac.

Please coordinate all of the longer tests with me so we don't waste resources unecessarily.

For all of the BOINC tests, please make sure to give me the workunit IDs.

I have a test matrix for the entire test plan, with 216 boxes to be filled in as each test is done. So far, 6 are green (complete). Only 210 more to go!

I've put the test matrix online in Google Drive. Clicking on that link will take you to the spreadsheet. Red boxes are available tests (Completed and validated always go together as a pair.) Please post here if you want to do any of the tests.
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74345 |

Honza

Volunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 15 Aug 05
Posts: 1931
ID: 352
Credit: 5,702,769,683
RAC: 1,041,236

Mike, will try to fill Windows AVX gap, at least for some tests.

Doing third part (manually running the new LLR against a known prime) is also possible.

Windows: AVX C B Size task Host Completed Validated Prime? 1 2 small PPSE 428118 385239008 Yes No 1 2 large SoB 428118 384279920 In progress 1 5 small SR5 428118 383935181 Yes No 1 5 large n/a 428118 -1 2 small SGS 428118 385000974 Yes No -1 2 large WOO 428118 383397081 In progress No -1 5 small SR5 428118 383876622 Yes No -1 5 large n/a 428118

EDIT: My Woodall WU got already validated yesterday so once my today task is finished, it should validate right away.
____________
My stats
Badge score: 1*1 + 5*1 + 8*3 + 9*11 + 10*1 + 11*1 + 12*3 = 186

ID: 74353 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

Mike, will try to fill Windows AVX gap, at least for some tests.

Doing third part (manually running the new LLR against a known prime) is also possible.

Windows: AVX C B Size task Host Completed Validated Prime? 1 2 small PPSE 428118 385239008 Yes No 1 2 large SoB 428118 1 5 small SR5 428118 383876622 In Progress 1 5 large n/a 428118 -1 2 small SGS 428118 -1 2 large WOO 428118 383397081 In progress -1 5 small SR5 428118 -1 5 large n/a 428118

EDIT: My Woodall WU got already validated yesterday so once my today task is finished, it should validate right away.

Honza,

I've marked you down for all of the AVX Windows BOINC and prime tests.

That SR5 is actually a Riesel (C == -1) test, so I moved it down to the last row. Fo the SR5s, you have to look in the "sr5_llr_##########" file in boinc/projects/www.primegrid.com to see if the number is +1 or -1. +1 will have a "P" in that file and -1 will have an "M". That file name is the same as the task name displayed in the BOINC task manager without the final "_#".

Sorry for neglecting to mention that.

Mike
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74359 |

Honza

Volunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 15 Aug 05
Posts: 1931
ID: 352
Credit: 5,702,769,683
RAC: 1,041,236

That SR5 is actually a Riesel (C == -1) test, so I moved it down to the last row. Fo the SR5s, you have to look in the "sr5_llr_##########" file in boinc/projects/www.primegrid.com to see if the number is +1 or -1. +1 will have a "P" in that file and -1 will have an "M". That file name is the same as the task name displayed in the BOINC task manager without the final "_#".

Table updated.

SR5, C=1 should be finished and veryfied in 2 hours, Woodall in 2 days.
SoB is in queue, it will take a while.
btw, FMA3 SoB is doing fine - should take "ony" 56 hours.

EDIT: another SR5 validated. I fell lucky that my SoB wingman seems reliable and validation shouldn't take 3+ months. Will finish SoB within a week.
____________
My stats
Badge score: 1*1 + 5*1 + 8*3 + 9*11 + 10*1 + 11*1 + 12*3 = 186

ID: 74369 |

JimB

Honorary cruncher
Send message
Joined: 4 Aug 11
Posts: 918
ID: 107307
Credit: 977,945,376
RAC: 45

ID: 74378 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

The manual prime tests will be granted PSA credit. The BOINC tests, since they're going through the normal BOINC system get normal BOINC credit.
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74389 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

GCW has been added as the small non-base 2 tests for both +1 and -1. Since this isn't a BOINC project, it's only a prime test.
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74390 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

For anyone wishing to do the prime tests (i.e., running the new LLR against a known prime) I've added a list of known primes for each test to the spreadsheet. The list is over on the right side and you may need to scroll the spreadsheet to see it.
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74391 |

Scott Brown

Volunteer moderator
Project administrator
Volunteer tester
Project scientist
Avatar

Send message
Joined: 17 Oct 05
Posts: 2329
ID: 1178
Credit: 15,582,675,723
RAC: 15,114,994

GCW has been added as the small non-base 2 tests for both +1 and -1. Since this isn't a BOINC project, it's only a prime test.

I can start these prime tests manually on a Win7 64-bit AVX machine and a Win Vista 32-bit machine this afternoon.

ID: 74397 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

GCW has been added as the small non-base 2 tests for both +1 and -1. Since this isn't a BOINC project, it's only a prime test.

I can start these prime tests manually on a Win7 64-bit AVX machine and a Win Vista 32-bit machine this afternoon.

I'll put you down for the AVX GCW prime tests. Your Vista 32 machine, is it a 3s bit cpu, or a 64 bit cpu? I'd rather run the 32 bit tests on real 32 bit machines. (I'm not sure that 32 CPUs were still being made when Vista came out.)
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74401 |

Honza

Volunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 15 Aug 05
Posts: 1931
ID: 352
Credit: 5,702,769,683
RAC: 1,041,236

Already running prime tests on Win 64-bit AVX CPU.
Have no 32-bit CPU/OS available at the moment so you are free to run these.

____________
My stats
Badge score: 1*1 + 5*1 + 8*3 + 9*11 + 10*1 + 11*1 + 12*3 = 186

ID: 74402 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

Already running prime tests on Win 64-bit AVX CPU.
Have no 32-bit CPU/OS available at the moment so you are free to run these.

If you're doing the GCWs, both you and Scott are doing the same tests. That's fine; their not very long. It would be a shame to duplicate an SoB test, however. :)
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74404 |

Scott Brown

Volunteer moderator
Project administrator
Volunteer tester
Project scientist
Avatar

Send message
Joined: 17 Oct 05
Posts: 2329
ID: 1178
Credit: 15,582,675,723
RAC: 15,114,994

GCW has been added as the small non-base 2 tests for both +1 and -1. Since this isn't a BOINC project, it's only a prime test.

I can start these prime tests manually on a Win7 64-bit AVX machine and a Win Vista 32-bit machine this afternoon.

I'll put you down for the AVX GCW prime tests. Your Vista 32 machine, is it a 3s bit cpu, or a 64 bit cpu? I'd rather run the 32 bit tests on real 32 bit machines. (I'm not sure that 32 CPUs were still being made when Vista came out.)

It is an old T2450 laptop, so indeed, a true 32-bit CPU.

I will be slightly delayed in getting these done as wind damage has knocked power out across campus where the machine is located and I am now home where I still have power. Will get them going tomorrow most likely.

ID: 74406 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

GCW has been added as the small non-base 2 tests for both +1 and -1. Since this isn't a BOINC project, it's only a prime test.

I can start these prime tests manually on a Win7 64-bit AVX machine and a Win Vista 32-bit machine this afternoon.

I'll put you down for the AVX GCW prime tests. Your Vista 32 machine, is it a 3s bit cpu, or a 64 bit cpu? I'd rather run the 32 bit tests on real 32 bit machines. (I'm not sure that 32 CPUs were still being made when Vista came out.)

It is an old T2450 laptop, so indeed, a true 32-bit CPU.

I will be slightly delayed in getting these done as wind damage has knocked power out across campus where the machine is located and I am now home where I still have power. Will get them going tomorrow most likely.

OK.

(And I'm glad you have power at home!)
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74408 |

Volunteer tester
Send message
Joined: 29 Sep 08
Posts: 87
ID: 29825
Credit: 365,054,285
RAC: 94,198

For anyone wishing to do the prime tests (i.e., running the new LLR against a known prime) I've added a list of known primes for each test to the spreadsheet. The list is over on the right side and you may need to scroll the spreadsheet to see it.

I have started the tests for 64-bit Linux/FMA3.

ID: 74411 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

For anyone wishing to do the prime tests (i.e., running the new LLR against a known prime) I've added a list of known primes for each test to the spreadsheet. The list is over on the right side and you may need to scroll the spreadsheet to see it.

I have started the tests for 64-bit Linux/FMA3.

Ok, please post the output from each test here as they complete. Thank you!
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74412 |

Honza

Volunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 15 Aug 05
Posts: 1931
ID: 352
Credit: 5,702,769,683
RAC: 1,041,236

There is a FMA3 SoB BOINC task validated
http://www.primegrid.com/workunit.php?wuid=378940805

Later today I should have all but SoB BOINC AVX tests done and validated; SoB should be by end of the week.

Manual primes on AVX so far:
2163*2^1255556+1 is prime! (377964 decimal digits) Time : 979.176 sec.
129897*68^129897+1 is prime! (238043 decimal digits) Time : 640.996 sec.
59912*5^1500861+1 is prime! (1049062 decimal digits) Time : 13965.236 sec.
1300274732577*2^1290000-1 is prime! (388341 decimal digits) Time : 1306.244 sec.
94112*121^94112-1 is prime! (196021 decimal digits, P = 9) Time : 4701.461 sec.

178658*5^1525224-1 is half done; 3752948*2^3752948-1 and 19249*2^13018586+1 are in queue.

____________
My stats
Badge score: 1*1 + 5*1 + 8*3 + 9*11 + 10*1 + 11*1 + 12*3 = 186

ID: 74424 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

There is a FMA3 SoB BOINC task validated
http://www.primegrid.com/workunit.php?wuid=378940805

Is that test with 3.8.12 or 3.8.13? I don't have WU IDs for the tasks you have in progress, so I have to go by the timestamps. It appears to me that this task was sent to you before we switched to 3.8.13. We need to test 3.8.13 specifically.

Manual primes on AVX so far:
2163*2^1255556+1 is prime! (377964 decimal digits) Time : 979.176 sec.
129897*68^129897+1 is prime! (238043 decimal digits) Time : 640.996 sec.
59912*5^1500861+1 is prime! (1049062 decimal digits) Time : 13965.236 sec.
1300274732577*2^1290000-1 is prime! (388341 decimal digits) Time : 1306.244 sec.
94112*121^94112-1 is prime! (196021 decimal digits, P = 9) Time : 4701.461 sec.

Could you post the full output from LLR please? I need to see everything, not just the "is prime" line. With all the software errors we've seen over the last few years I'm being very picky with the testing on software right now.

(By "full output", I mean everything EXCEPT the line with the "Iteration 10000... Iteration 20000..." counter. I don't need a million lines of that. Everything else, yes. On Linux or Mac, that includes doing a run for a few seconds at each number tested redirecting stdout into a file so I can see the transform line. On those OSs that gets overwritten and you can't see it visually. This isn't a problem on Windows.)
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74427 |

Scott Brown

Volunteer moderator
Project administrator
Volunteer tester
Project scientist
Avatar

Send message
Joined: 17 Oct 05
Posts: 2329
ID: 1178
Credit: 15,582,675,723
RAC: 15,114,994

AVX Tests (i7-2670QM 64-bit Win7 Home Premium):

C:\Users\sbrow\Downloads\prpclient-5.2.8-windows>cllr64.exe -d -q"129897*68^129897+1" Base factorized as : 2^2*17 Base prime factor(s) taken : 17 Starting N-1 prime test of 129897*68^129897+1 Using zero-padded AVX FFT length 100K, Pass1=320, Pass2=320, a = 3 129897*68^129897+1 may be prime, trying to compute gcd's 3^((N-1)/17)-1 is coprime to N! 129897*68^129897+1 is prime! (238043 decimal digits) Time : 960.809 sec.

C:\Users\sbrow\Downloads\prpclient-5.2.8-windows>cllr64.exe -d -q"94112*121^9411 2-1" Base prime factor(s) taken : 11 Starting N+1 prime test of 94112*121^94112-1 Using zero-padded AVX FFT length 84K, Pass1=448, Pass2=192, a = 3 94112*121^94112-1 may be prime. Starting Lucas sequence... Using zero-padded AVX FFT length 84K, Pass1=448, Pass2=192, P = 4 94112*121^94112-1 may be prime, trying to compute gcd's Restarting Lucas sequence with P = 8 Using zero-padded AVX FFT length 84K, Pass1=448, Pass2=192, P = 8 94112*121^94112-1 may be prime, trying to compute gcd's Restarting Lucas sequence with P = 9 Using zero-padded AVX FFT length 84K, Pass1=448, Pass2=192, P = 9 94112*121^94112-1 may be prime, trying to compute gcd's U((N+1)/11) is coprime to N! 94112*121^94112-1 is prime! (196021 decimal digits, P = 9) Time : 7516.738 sec.

ID: 74428 |

Scott Brown

Volunteer moderator
Project administrator
Volunteer tester
Project scientist
Avatar

Send message
Joined: 17 Oct 05
Posts: 2329
ID: 1178
Credit: 15,582,675,723
RAC: 15,114,994

32-bit machine crapped out on me, so I had to do a repair. It is back up and testing the +1 GCW right now. I am also running the PPSE manual prime test on the second core. I'll go ahead and sign the box up for all of the 32-bit manual prime test in Windows...that is if we really want a true 32-bit CPU test for all of these. I am thinking that the SoB will take longer than the deadlines under BOINC such that only a faster processor (i.e., those made well after 64-bit CPUs were standard) could complete it.

Even without SoB, some of these will take a while. But at least it is a dual-core 32-bit CPU so I can do two at a time. :)

ID: 74429 |

Honza

Volunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 15 Aug 05
Posts: 1931
ID: 352
Credit: 5,702,769,683
RAC: 1,041,236

Is that test with 3.8.12 or 3.8.13? I don't have WU IDs for the tasks you have in progress, so I have to go by the timestamps. It appears to me that this task was sent to you before we switched to 3.8.13. We need to test 3.8.13 specifically.

Yes, it was just before 3.8.13 fixed-belease was available.
It was when I asked if it is safe and only non-base 2 (SR5) is potencionally affected.
This is also not SB or IB CPU but Haswell.

Could you post the full output from LLR please? I need to see everything, not just the "is prime" line. With all the software errors we've seen over the last few years I'm being very picky with the testing on software right now.

CMD window is still open so here we go...

c:\_Honza>cllr64.exe -q"2163*2^1255556+1" -d Starting Proth prime test of 2163*2^1255556+1 Using all-complex AVX FFT length 96K, Pass1=128, Pass2=768, a = 13 2163*2^1255556+1 is prime! (377964 decimal digits) Time : 979.176 sec. c:\_Honza>cllr64.exe -q"129897*68^129897+1" -d Base factorized as : 2^2*17 Base prime factor(s) taken : 17 Starting N-1 prime test of 129897*68^129897+1 Using zero-padded AVX FFT length 100K, Pass1=320, Pass2=320, a = 3 129897*68^129897+1 may be prime, trying to compute gcd's 3^((N-1)/17)-1 is coprime to N! 129897*68^129897+1 is prime! (238043 decimal digits) Time : 640.996 sec. c:\_Honza>cllr64.exe -q"59912*5^1500861+1" -d Base prime factor(s) taken : 5 Starting N-1 prime test of 59912*5^1500861+1 Using all-complex AVX FFT length 384K, Pass1=384, Pass2=1K, a = 3 59912*5^1500861+1 may be prime, trying to compute gcd's 3^((N-1)/5)-1 is coprime to N! 59912*5^1500861+1 is prime! (1049062 decimal digits) Time : 13965.236 sec. c:\_Honza>cllr64.exe -q"1300274732577*2^1290000-1" -d Starting Lucas Lehmer Riesel prime test of 1300274732577*2^1290000-1 Using zero-padded AVX FFT length 128K, Pass1=512, Pass2=256 V1 = 5 ; Computing U0...done. 1300274732577*2^1290000-1 is prime! (388341 decimal digits) Time : 1306.244 sec. c:\_Honza>cllr64.exe -q"94112*121^94112-1" -d Base prime factor(s) taken : 11 Starting N+1 prime test of 94112*121^94112-1 Using zero-padded AVX FFT length 84K, Pass1=448, Pass2=192, a = 3 94112*121^94112-1 may be prime. Starting Lucas sequence... Using zero-padded AVX FFT length 84K, Pass1=448, Pass2=192, P = 4 94112*121^94112-1 may be prime, trying to compute gcd's Restarting Lucas sequence with P = 8 Using zero-padded AVX FFT length 84K, Pass1=448, Pass2=192, P = 8 94112*121^94112-1 may be prime, trying to compute gcd's Restarting Lucas sequence with P = 9 Using zero-padded AVX FFT length 84K, Pass1=448, Pass2=192, P = 9 94112*121^94112-1 may be prime, trying to compute gcd's U((N+1)/11) is coprime to N! 94112*121^94112-1 is prime! (196021 decimal digits, P = 9) Time : 4701.461 sec. c:\_Honza>cllr64.exe -q"178658*5^1525224-1" -d Base prime factor(s) taken : 5 Starting N+1 prime test of 178658*5^1525224-1 Using AVX FFT length 384K, Pass1=384, Pass2=1K, a = 3 178658*5^1525224-1 may be prime. Starting Lucas sequence... Using AVX FFT length 384K, Pass1=384, Pass2=1K, P = 5 178658*5^1525224-1, bit: 2500000 / 3541476 [70.59%]. Time per bit: 19.578 ms.

____________
My stats
Badge score: 1*1 + 5*1 + 8*3 + 9*11 + 10*1 + 11*1 + 12*3 = 186

ID: 74430 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

32-bit machine crapped out on me, so I had to do a repair. It is back up and testing the +1 GCW right now. I am also running the PPSE manual prime test on the second core. I'll go ahead and sign the box up for all of the 32-bit manual prime test in Windows...that is if we really want a true 32-bit CPU test for all of these. I am thinking that the SoB will take longer than the deadlines under BOINC such that only a faster processor (i.e., those made well after 64-bit CPUs were standard) could complete it.

Even without SoB, some of these will take a while. But at least it is a dual-core 32-bit CPU so I can do two at a time. :)

Yes, it will take a while, and yes, it won't make the deadline. Let me worry about that. :)

I suspect that running both cores will impact the execution speed because of memory contention, the same as it does on newer CPUs. Fortunately, my 32 bit CPU is an older one with only one core so I don't have THAT problem. (Yes, that's a joke. :) )

I've you marked down for all the Win32 manual prime tests, and recorded your two manual tests in Win64 AVX GCW. Thanks!
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74432 |

Stojag

Send message
Joined: 8 Jun 10
Posts: 35
ID: 62023
Credit: 42,138,433
RAC: 0

I can do some FMA3 validation tests for Win64

As i'm fairly new to the comunity, let me know if i'm not doing correctly.
I'm using an Haswell Xeon E3-1230V3 @3.7GHz turbo with 2 x 8GB DIMMs at 1600MHz
At the moment i'm just running this one computation on my CPU

Here is the output from the 3.8.13 Win64 version:

C:\Users\Christof\Downloads>cllr64 -d -q"2163*2^1255556+1" Starting Proth prime test of 2163*2^1255556+1 Using all-complex FMA3 FFT length 96K, Pass1=128, Pass2=768, a = 13 2163*2^1255556+1 is prime! (377964 decimal digits) Time : 467.694 sec.

At the moment i'm running the SoB 19249*2^13018586+1 task, that may take another 24h to finish

ID: 74433 |

Dave

Send message
Joined: 13 Feb 12
Posts: 3062
ID: 130544
Credit: 2,114,436,724
RAC: 1,422,811

I did a SoB on a 2.5 Celeron (machine ID 256349 if you're interested) some time ago + it took 23 days.

ID: 74434 |

Volunteer tester
Send message
Joined: 29 Sep 08
Posts: 87
ID: 29825
Credit: 365,054,285
RAC: 94,198

I have started the tests for 64-bit Linux/FMA3.

Ok, please post the output from each test here as they complete. Thank you!

Here it is.

Starting Proth prime test of 2163*2^1255556+1 2163*2^1255556+1 is prime! (377964 decimal digits) Time : 438.901 sec. Starting Proth prime test of 19249*2^13018586+1 19249*2^13018586+1 is prime! (3918990 decimal digits) Time : 72960.027 sec. Base factorized as : 2^2*17 Base prime factor(s) taken : 17 Starting N-1 prime test of 129897*68^129897+1 129897*68^129897+1 may be prime, trying to compute gcd's 3^((N-1)/17)-1 is coprime to N! 129897*68^129897+1 is prime! (238043 decimal digits) Time : 303.968 sec. Base prime factor(s) taken : 5 Starting N-1 prime test of 59912*5^1500861+1 59912*5^1500861+1 may be prime, trying to compute gcd's 3^((N-1)/5)-1 is coprime to N! 59912*5^1500861+1 is prime! (1049062 decimal digits) Time : 6657.758 sec. Starting Lucas Lehmer Riesel prime test of 1300274732577*2^1290000-1 Using zero-padded FMA3 FFT length 128K, Pass1=512, Pass2=256 1300274732577*2^1290000-1 is prime! (388341 decimal digits) Time : 613.095 sec. Starting Lucas Lehmer Riesel prime test of 3752948*2^3752948-1 Using zero-padded FMA3 FFT length 384K, Pass1=256, Pass2=1536 3752948*2^3752948-1 is prime! (1129757 decimal digits) Time : 6681.019 sec. Base prime factor(s) taken : 11 Starting N+1 prime test of 94112*121^94112-1 94112*121^94112-1 may be prime. Starting Lucas sequence... 94112*121^94112-1 may be prime, trying to compute gcd's Restarting Lucas sequence with P = 8 94112*121^94112-1 may be prime, trying to compute gcd's Restarting Lucas sequence with P = 9 94112*121^94112-1 may be prime, trying to compute gcd's U((N+1)/11) is coprime to N! 94112*121^94112-1 is prime! (196021 decimal digits, P = 9) Time : 2002.629 sec. Base prime factor(s) taken : 5 Starting N+1 prime test of 178658*5^1525224-1 178658*5^1525224-1 may be prime. Starting Lucas sequence... 178658*5^1525224-1 may be prime, trying to compute gcd's Restarting Lucas sequence with P = 11 178658*5^1525224-1 may be prime, trying to compute gcd's U((N+1)/5) is coprime to N! 178658*5^1525224-1 is prime! (1066092 decimal digits, P = 11) Time : 53684.683 sec.

ID: 74440 |

Honza

Volunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 15 Aug 05
Posts: 1931
ID: 352
Credit: 5,702,769,683
RAC: 1,041,236

BOINC AVX only SoB in progress, rest is completed and validated.

Windows: AVX C B Size task Host Completed Validated Prime? 1 2 small PPSE 428118 385239008 Yes No 1 2 large SoB 428118 384279920 In progress 1 5 small SR5 428118 383935181 Yes No 1 5 large n/a 428118 -1 2 small SGS 428118 385000974 Yes No -1 2 large WOO 428118 383397081 Yes No -1 5 small SR5 428118 383876622 Yes No -1 5 large n/a 428118

____________
My stats
Badge score: 1*1 + 5*1 + 8*3 + 9*11 + 10*1 + 11*1 + 12*3 = 186

ID: 74442 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

Is that test with 3.8.12 or 3.8.13? I don't have WU IDs for the tasks you have in progress, so I have to go by the timestamps. It appears to me that this task was sent to you before we switched to 3.8.13. We need to test 3.8.13 specifically.

Yes, it was just before 3.8.13 fixed-belease was available.
It was when I asked if it is safe and only non-base 2 (SR5) is potencionally affected.
This is also not SB or IB CPU but Haswell.

These tests are to validate the 3.8.13 executables. Tests performed on 3.8.12 (or anything except 3.8.12) don't help, sorry.

I recorded all the primes, thank you.

____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74443 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

I can do some FMA3 validation tests for Win64

As i'm fairly new to the comunity, let me know if i'm not doing correctly.
I'm using an Haswell Xeon E3-1230V3 @3.7GHz turbo with 2 x 8GB DIMMs at 1600MHz
At the moment i'm just running this one computation on my CPU

Here is the output from the 3.8.13 Win64 version:

C:\Users\Christof\Downloads>cllr64 -d -q"2163*2^1255556+1" Starting Proth prime test of 2163*2^1255556+1 Using all-complex FMA3 FFT length 96K, Pass1=128, Pass2=768, a = 13 2163*2^1255556+1 is prime! (377964 decimal digits) Time : 467.694 sec.

At the moment i'm running the SoB 19249*2^13018586+1 task, that may take another 24h to finish

Thank you!

If possible, try to let me know that you're starting a test so that I can mark them as reserved/in-progress on the spreadsheet so nobody needlessly duplicates the work.

____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74444 |

Honza

Volunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 15 Aug 05
Posts: 1931
ID: 352
Credit: 5,702,769,683
RAC: 1,041,236

These tests are to validate the 3.8.13 executables. Tests performed on 3.8.12 (or anything except 3.8.12) don't help, sorry.

OK, understood.
There are 3 other SoB tasks started today...
____________
My stats
Badge score: 1*1 + 5*1 + 8*3 + 9*11 + 10*1 + 11*1 + 12*3 = 186

ID: 74445 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

I have started the tests for 64-bit Linux/FMA3.

Ok, please post the output from each test here as they complete. Thank you!

Here it is.

Starting Proth prime test of 2163*2^1255556+1 2163*2^1255556+1 is prime! (377964 decimal digits) Time : 438.901 sec. Starting Proth prime test of 19249*2^13018586+1 19249*2^13018586+1 is prime! (3918990 decimal digits) Time : 72960.027 sec. Base factorized as : 2^2*17 Base prime factor(s) taken : 17 Starting N-1 prime test of 129897*68^129897+1 129897*68^129897+1 may be prime, trying to compute gcd's 3^((N-1)/17)-1 is coprime to N! 129897*68^129897+1 is prime! (238043 decimal digits) Time : 303.968 sec. Base prime factor(s) taken : 5 Starting N-1 prime test of 59912*5^1500861+1 59912*5^1500861+1 may be prime, trying to compute gcd's 3^((N-1)/5)-1 is coprime to N! 59912*5^1500861+1 is prime! (1049062 decimal digits) Time : 6657.758 sec. Starting Lucas Lehmer Riesel prime test of 1300274732577*2^1290000-1 Using zero-padded FMA3 FFT length 128K, Pass1=512, Pass2=256 1300274732577*2^1290000-1 is prime! (388341 decimal digits) Time : 613.095 sec. Starting Lucas Lehmer Riesel prime test of 3752948*2^3752948-1 Using zero-padded FMA3 FFT length 384K, Pass1=256, Pass2=1536 3752948*2^3752948-1 is prime! (1129757 decimal digits) Time : 6681.019 sec. Base prime factor(s) taken : 11 Starting N+1 prime test of 94112*121^94112-1 94112*121^94112-1 may be prime. Starting Lucas sequence... 94112*121^94112-1 may be prime, trying to compute gcd's Restarting Lucas sequence with P = 8 94112*121^94112-1 may be prime, trying to compute gcd's Restarting Lucas sequence with P = 9 94112*121^94112-1 may be prime, trying to compute gcd's U((N+1)/11) is coprime to N! 94112*121^94112-1 is prime! (196021 decimal digits, P = 9) Time : 2002.629 sec. Base prime factor(s) taken : 5 Starting N+1 prime test of 178658*5^1525224-1 178658*5^1525224-1 may be prime. Starting Lucas sequence... 178658*5^1525224-1 may be prime, trying to compute gcd's Restarting Lucas sequence with P = 11 178658*5^1525224-1 may be prime, trying to compute gcd's U((N+1)/5) is coprime to N! 178658*5^1525224-1 is prime! (1066092 decimal digits, P = 11) Time : 53684.683 sec.

Thank you!
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74446 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

BOINC AVX only SoB in progress, rest is completed and validated.

Windows: AVX C B Size task Host Completed Validated Prime? 1 2 small PPSE 428118 385239008 Yes No 1 2 large SoB 428118 384279920 In progress 1 5 small SR5 428118 383935181 Yes No 1 5 large n/a 428118 -1 2 small SGS 428118 385000974 Yes No -1 2 large WOO 428118 383397081 Yes No -1 5 small SR5 428118 383876622 Yes No -1 5 large n/a 428118

Thank you!
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74447 |

Scott Brown

Volunteer moderator
Project administrator
Volunteer tester
Project scientist
Avatar

Send message
Joined: 17 Oct 05
Posts: 2329
ID: 1178
Credit: 15,582,675,723
RAC: 15,114,994

32-bit GCW +1 test using T2450 on Win Vista Home Premium:

C:\Users\Scott\Downloads>cllr.exe -d -q"129897*68^129897+1" Base factorized as : 2^2*17 Base prime factor(s) taken : 17 Starting N-1 prime test of 129897*68^129897+1 Using zero-padded FFT length 112K, Pass1=448, Pass2=256, a = 3 129897*68^129897+1 may be prime, trying to compute gcd's 3^((N-1)/17)-1 is coprime to N! 129897*68^129897+1 is prime! (238043 decimal digits) Time : 4958.104 sec.

ID: 74457 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

32-bit GCW +1 test using T2450 on Win Vista Home Premium:

C:\Users\Scott\Downloads>cllr.exe -d -q"129897*68^129897+1" Base factorized as : 2^2*17 Base prime factor(s) taken : 17 Starting N-1 prime test of 129897*68^129897+1 Using zero-padded FFT length 112K, Pass1=448, Pass2=256, a = 3 129897*68^129897+1 may be prime, trying to compute gcd's 3^((N-1)/17)-1 is coprime to N! 129897*68^129897+1 is prime! (238043 decimal digits) Time : 4958.104 sec.

Thank you!
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74459 |

Scott Brown

Volunteer moderator
Project administrator
Volunteer tester
Project scientist
Avatar

Send message
Joined: 17 Oct 05
Posts: 2329
ID: 1178
Credit: 15,582,675,723
RAC: 15,114,994

32-bit PPSE (Pentium M 2.1Ghz on Win7 Pro):

C:\Users\sbrow\Downloads\cllr38d>cllr.exe -d -q"2163*2^1255556+1" Starting Proth prime test of 2163*2^1255556+1 Using all-complex FFT length 96K, Pass1=384, Pass2=256, a = 13 2163*2^1255556+1 is prime! (377964 decimal digits) Time : 6448.499 sec.

ID: 74460 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

32-bit PPSE (Pentium M 2.1Ghz on Win7 Pro):

C:\Users\sbrow\Downloads\cllr38d>cllr.exe -d -q"2163*2^1255556+1" Starting Proth prime test of 2163*2^1255556+1 Using all-complex FFT length 96K, Pass1=384, Pass2=256, a = 13 2163*2^1255556+1 is prime! (377964 decimal digits) Time : 6448.499 sec.

Got it, thanks!
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74461 |

Stojag

Send message
Joined: 8 Jun 10
Posts: 35
ID: 62023
Credit: 42,138,433
RAC: 0

Result for SoB FMA3 Windows 64bit

C:\Users\Christof\Downloads>cllr64 -d -q"19249*2^13018586+1" Starting Proth prime test of 19249*2^13018586+1 Using all-complex FMA3 FFT length 1152K, Pass1=768, Pass2=1536, a = 3 19249*2^13018586+1 is prime! (3918990 decimal digits) Time : 67527.619 sec.

I'll do the 6 remaining prime test now

ID: 74463 |

Gary Craig
Volunteer tester
Avatar

Send message
Joined: 30 Dec 09
Posts: 3213
ID: 52890
Credit: 1,005,618,748
RAC: 0

I decided to grab some low-hanging fruit... here are the prime tests for the four "short" tests for Mac SSE3 (an iMac, C2D) on 3.8.13. I edited out the "heartbeat" progress-update lines. In the last test, I missed catching one of the "Using..." messages due to the overwrite problem. In the future for anything longer, I'll coordinate first... these were short enough I figured it wouldn't matter.

./llr64 -d -q"2163*2^1255556+1" Starting Proth prime test of 2163*2^1255556+1 Using all-complex FFT length 96K, Pass1=128, Pass2=768, a = 13 2163*2^1255556+1 is prime! (377964 decimal digits) Time : 1653.845 sec. ./llr64 -d -q"129897*68^129897+1" Base factorized as : 2^2*17 Base prime factor(s) taken : 17 Starting N-1 prime test of 129897*68^129897+1 Using zero-padded FFT length 112K, Pass1=448, Pass2=256, a = 3 129897*68^129897+1 may be prime, trying to compute gcd's 3^((N-1)/17)-1 is coprime to N! 129897*68^129897+1 is prime! (238043 decimal digits) Time : 1756.204 sec. ./llr64 -d -q"1300274732577*2^1290000-1" Starting Lucas Lehmer Riesel prime test of 1300274732577*2^1290000-1 Using zero-padded FFT length 128K, Pass1=128, Pass2=1K V1 = 5 ; Computing U0...done.Starting Lucas-Lehmer loop... 1300274732577*2^1290000-1 is prime! (388341 decimal digits) Time : 2182.139 sec. ./llr64 -d -q"94112*121^94112-1" Base prime factor(s) taken : 11 Starting N+1 prime test of 94112*121^94112-1 Using zero-padded type-1 FFT length 84K, Pass1=112, Pass2=768, a = 3 94112*121^94112-1 may be prime. Starting Lucas sequence... 94112*121^94112-1 may be prime, trying to compute gcd's Restarting Lucas sequence with P = 8 94112*121^94112-1 may be prime, trying to compute gcd's Restarting Lucas sequence with P = 9 Using zero-padded type-1 FFT length 84K, Pass1=112, Pass2=768, P = 9 94112*121^94112-1 may be prime, trying to compute gcd's U((N+1)/11) is coprime to N! 94112*121^94112-1 is prime! (196021 decimal digits, P = 9) Time : 9766.210 sec.

I'll have to defer anything else here until next week, so don't reserve anything for me at this point.
--Gary

ID: 74464 |

Honza

Volunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 15 Aug 05
Posts: 1931
ID: 352
Credit: 5,702,769,683
RAC: 1,041,236

One more and one left to go on manual AVX.

c:\_Honza>cllr64.exe -q"3752948*2^3752948-1" -d Starting Lucas Lehmer Riesel prime test of 3752948*2^3752948-1 Using zero-padded AVX FFT length 384K, Pass1=384, Pass2=1K V1 = 4 ; Computing U0...done. 3752948*2^3752948-1 is prime! (1129757 decimal digits) Time : 12407.249 sec.

____________
My stats
Badge score: 1*1 + 5*1 + 8*3 + 9*11 + 10*1 + 11*1 + 12*3 = 186

ID: 74467 |

Scott Brown

Volunteer moderator
Project administrator
Volunteer tester
Project scientist
Avatar

Send message
Joined: 17 Oct 05
Posts: 2329
ID: 1178
Credit: 15,582,675,723
RAC: 15,114,994

32-bit SGS (Pentium M 2.1Ghz on Win7 Pro):

C:\Users\sbrow\Downloads\cllr38d>cllr.exe -d -q"1300274732577*2^1290000-1" Starting Lucas Lehmer Riesel prime test of 1300274732577*2^1290000-1 Using zero-padded FFT length 128K, Pass1=512, Pass2=256 V1 = 5 ; Computing U0...done. 1300274732577*2^1290000-1 is prime! (388341 decimal digits) Time : 8966.425 sec.

ID: 74468 |

Stojag

Send message
Joined: 8 Jun 10
Posts: 35
ID: 62023
Credit: 42,138,433
RAC: 0

completet some more FMA3 Win64 prime tests

C:\Users\Christof\Downloads>cllr64 -d -q"129897*68^129897+1" Base factorized as : 2^2*17 Base prime factor(s) taken : 17 Starting N-1 prime test of 129897*68^129897+1 Using zero-padded FMA3 FFT length 100K, Pass1=320, Pass2=320, a = 3 129897*68^129897+1 may be prime, trying to compute gcd's 3^((N-1)/17)-1 is coprime to N! 129897*68^129897+1 is prime! (238043 decimal digits) Time : 310.804 sec. C:\Users\Christof\Downloads>cllr64 -d -q"59912*5^1500861+1" Base prime factor(s) taken : 5 Starting N-1 prime test of 59912*5^1500861+1 Using all-complex FMA3 FFT length 384K, Pass1=512, Pass2=768, a = 3 59912*5^1500861+1 may be prime, trying to compute gcd's 3^((N-1)/5)-1 is coprime to N! 59912*5^1500861+1 is prime! (1049062 decimal digits) Time : 5844.303 sec. C:\Users\Christof\Downloads>cllr64 -d -q"1300274732577*2^1290000-1" Starting Lucas Lehmer Riesel prime test of 1300274732577*2^1290000-1 Using zero-padded FMA3 FFT length 128K, Pass1=512, Pass2=256 V1 = 5 ; Computing U0...done. 1300274732577*2^1290000-1 is prime! (388341 decimal digits) Time : 631.181 sec. C:\Users\Christof\Downloads>cllr64 -d -q"3752948*2^3752948-1" Starting Lucas Lehmer Riesel prime test of 3752948*2^3752948-1 Using zero-padded FMA3 FFT length 384K, Pass1=256, Pass2=1536 V1 = 4 ; Computing U0...done. 3752948*2^3752948-1 is prime! (1129757 decimal digits) Time : 5745.989 sec. C:\Users\Christof\Downloads>cllr64 -d -q"94112*121^94112-1" Base prime factor(s) taken : 11 Starting N+1 prime test of 94112*121^94112-1 Using zero-padded FMA3 FFT length 84K, Pass1=448, Pass2=192, a = 3 94112*121^94112-1 may be prime. Starting Lucas sequence... Using zero-padded FMA3 FFT length 84K, Pass1=448, Pass2=192, P = 4 94112*121^94112-1 may be prime, trying to compute gcd's Restarting Lucas sequence with P = 8 Using zero-padded FMA3 FFT length 84K, Pass1=448, Pass2=192, P = 8 94112*121^94112-1 may be prime, trying to compute gcd's Restarting Lucas sequence with P = 9 Using zero-padded FMA3 FFT length 84K, Pass1=448, Pass2=192, P = 9 94112*121^94112-1 may be prime, trying to compute gcd's U((N+1)/11) is coprime to N! 94112*121^94112-1 is prime! (196021 decimal digits, P = 9) Time : 2039.150 sec.

Right now i'm running the last one (SR5)

ID: 74469 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

Result for SoB FMA3 Windows 64bit

C:\Users\Christof\Downloads>cllr64 -d -q"19249*2^13018586+1" Starting Proth prime test of 19249*2^13018586+1 Using all-complex FMA3 FFT length 1152K, Pass1=768, Pass2=1536, a = 3 19249*2^13018586+1 is prime! (3918990 decimal digits) Time : 67527.619 sec.

I'll do the 6 remaining prime test now

Thanks!
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74470 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

I decided to grab some low-hanging fruit... here are the prime tests for the four "short" tests for Mac SSE3 (an iMac, C2D) on 3.8.13. I edited out the "heartbeat" progress-update lines. In the last test, I missed catching one of the "Using..." messages due to the overwrite problem. In the future for anything longer, I'll coordinate first... these were short enough I figured it wouldn't matter.

I'll have to defer anything else here until next week, so don't reserve anything for me at this point.
--Gary

Thanks!
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74471 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

One more and one left to go on manual AVX.

c:\_Honza>cllr64.exe -q"3752948*2^3752948-1" -d Starting Lucas Lehmer Riesel prime test of 3752948*2^3752948-1 Using zero-padded AVX FFT length 384K, Pass1=384, Pass2=1K V1 = 4 ; Computing U0...done. 3752948*2^3752948-1 is prime! (1129757 decimal digits) Time : 12407.249 sec.

Got that one, thanks. What happened to the SR5 -5? Wasn't that test halfway done in your last post?

Ok, I see what happened. It's marked complete in the very last post which was just a table, but the last output you posted was when it was still half done. Did it finish with a prime result?

Other that that, you're just working on the SoB, correct?
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74472 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

completet some more FMA3 Win64 prime tests

Right now i'm running the last one (SR5)

Thank you!
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74473 |

Honza

Volunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 15 Aug 05
Posts: 1931
ID: 352
Credit: 5,702,769,683
RAC: 1,041,236

Got that one, thanks. What happened to the SR5 -5? Wasn't that test halfway done in your last post?

Ok, I see what happened. It's marked complete in the very last post which was just a table, but the last output you posted was when it was still half done. Did it finish with a prime result?

Other that that, you're just working on the SoB, correct?

Yeah, about that SR5 -1.
That one was resumed later on.

And yes, there are only BOINC SoB (half-done) and manual SoB left on my AVX.
And BOINC FMA3 SoBs.

c:\_Honza>cllr64.exe -q"178658*5^1525224-1" -d Base prime factor(s) taken : 5 Resuming Lucas sequence at bit 28467 [0.80%] Using AVX FFT length 384K, Pass1=384, Pass2=1K, P = 11 178658*5^1525224-1 may be prime, trying to compute gcd's U((N+1)/5) is coprime to N! 178658*5^1525224-1 is prime! (1066092 decimal digits, P = 11) Time : 127836.780 sec.

____________
My stats
Badge score: 1*1 + 5*1 + 8*3 + 9*11 + 10*1 + 11*1 + 12*3 = 186

ID: 74474 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

And yes, there are only BOINC SoB (half-done) and manual SoB left on my AVX.
And BOINC FMA3 SoBs.

I didn't know you were doing any FMA3. Is that on Linux? If so, what tests are you running and what are the WU ids? All the Windows and Linux FMA3 work is already complete or reserved. I'll do the linux FMA3 once I finish the Windows tests, but if someone with a native Linux Haswell can run those it will save me the trouble of setting up the VM. (I'll have to do that eventually anyway, so it's not a big deal.)
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74476 |

Scott Brown

Volunteer moderator
Project administrator
Volunteer tester
Project scientist
Avatar

Send message
Joined: 17 Oct 05
Posts: 2329
ID: 1178
Credit: 15,582,675,723
RAC: 15,114,994

32-bit GCW -1 test using T2450 on Win Vista Home Premium:

C:\Users\Scott\Downloads>cllr.exe -d -q"94112*121^94112-1" Base prime factor(s) taken : 11 Starting N+1 prime test of 94112*121^94112-1 Using zero-padded type-1 FFT length 84K, Pass1=112, Pass2=768, a = 3 94112*121^94112-1 may be prime. Starting Lucas sequence... Using zero-padded type-1 FFT length 84K, Pass1=112, Pass2=768, P = 4 94112*121^94112-1 may be prime, trying to compute gcd's Restarting Lucas sequence with P = 8 Using zero-padded type-1 FFT length 84K, Pass1=112, Pass2=768, P = 8 94112*121^94112-1 may be prime, trying to compute gcd's Restarting Lucas sequence with P = 9 Using zero-padded type-1 FFT length 84K, Pass1=112, Pass2=768, P = 9 94112*121^94112-1 may be prime, trying to compute gcd's U((N+1)/11) is coprime to N! 94112*121^94112-1 is prime! (196021 decimal digits, P = 9) Time : 26847.351 sec.

ID: 74478 |

Honza

Volunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 15 Aug 05
Posts: 1931
ID: 352
Credit: 5,702,769,683
RAC: 1,041,236

I didn't know you were doing any FMA3. Is that on Linux? If so, what tests are you running and what are the WU ids? All the Windows and Linux FMA3 work is already complete or reserved. I'll do the linux FMA3 once I finish the Windows tests, but if someone with a native Linux Haswell can run those it will save me the trouble of setting up the VM. (I'll have to do that eventually anyway, so it's not a big deal.)

This is my regular BOINC host 376156 runing Win7 x64.
WU IDs are 384279967, 384279979 and 384279952.
It finished SoB using 3.8.12 but we need 3.8.13.
Machine would run SoB anyway and will be monitored closely.
I guess no need to put in the test plan...
____________
My stats
Badge score: 1*1 + 5*1 + 8*3 + 9*11 + 10*1 + 11*1 + 12*3 = 186

ID: 74479 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

32-bit GCW -1 test using T2450 on Win Vista Home Premium:

Thanks!

____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74481 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

I didn't know you were doing any FMA3. Is that on Linux? If so, what tests are you running and what are the WU ids? All the Windows and Linux FMA3 work is already complete or reserved. I'll do the linux FMA3 once I finish the Windows tests, but if someone with a native Linux Haswell can run those it will save me the trouble of setting up the VM. (I'll have to do that eventually anyway, so it's not a big deal.)

This is my regular BOINC host 376156 runing Win7 x64.
WU IDs are 384279967, 384279979 and 384279952.
It finished SoB using 3.8.12 but we need 3.8.13.
Machine would run SoB anyway and will be monitored closely.
I guess no need to put in the test plan...

Ok, no problem.
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74482 |

Prime Al Screen

Send message
Joined: 1 Dec 09
Posts: 240
ID: 50942
Credit: 38,208,946
RAC: 0

FMA3 tasks really, really run extremely hot! Compared to current BOINC app I get about 10oC increase in CPU temps and thermal warnings. This is possibly due to turbo boosting and the known cooling issues of these cpus. So I would not recommend this FMA3 version.

Linux 64bit i7-4770K CPU @ 3.50GHz host using the downloaded static version for the short tests. I will do the 2 SR5 tests but the remaining 2 will depend on challenge progress.

Starting Proth prime test of 2163*2^1255556+1 2163*2^1255556+1 is prime! (377964 decimal digits) Time : 387.274 sec.

Base factorized as : 2^2*17 Base prime factor(s) taken : 17 Starting N-1 prime test of 129897*68^129897+1 129897*68^129897+1 may be prime, trying to compute gcd's 3^((N-1)/17)-1 is coprime to N! 129897*68^129897+1 is prime! (238043 decimal digits) Time : 262.580 sec.

Starting Lucas Lehmer Riesel prime test of 1300274732577*2^1290000-1 Using zero-padded FMA3 FFT length 128K, Pass1=512, Pass2=256 1300274732577*2^1290000-1 is prime! (388341 decimal digits) Time : 540.096 sec.

Base prime factor(s) taken : 11 Starting N+1 prime test of 94112*121^94112-1 94112*121^94112-1 may be prime. Starting Lucas sequence... 94112*121^94112-1 may be prime, trying to compute gcd's Restarting Lucas sequence with P = 8 94112*121^94112-1 may be prime, trying to compute gcd's Restarting Lucas sequence with P = 9 94112*121^94112-1 may be prime, trying to compute gcd's U((N+1)/11) is coprime to N! 94112*121^94112-1 is prime! (196021 decimal digits, P = 9) Time : 1827.015 sec.

ID: 74483 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

FMA3 tasks really, really run extremely hot! Compared to current BOINC app I get about 10oC increase in CPU temps and thermal warnings. This is possibly due to turbo boosting and the known cooling issues of these cpus. So I would not recommend this FMA3 version.

The increase in speed comes with an increase in power, and that in turn increases thermal output (since 100% of the electricity consumed is turned into heat). As always, our warning "Make sure your system is up to it" applies.

If you're running with normal speed memory (up to 1600 MHz) the stock cooler should be able to handle the load because the the memory is a bottleneck and will slow down the CPU. But if you're using faster memory, even if you're not overclocking the CPU, the stock cooler probably can't cool the CPU sufficiently and the CPU will slow down to lower the temperatures. Also, the smaller the numbers, the hotter the CPU will run.

As you mentioned, *some* Haswell chips have been known to have cooling problems due to the thermal compound used between the chip itself and the heat spreader (the metal case surrounding the chip). Available options are to RMA the CPU, run it at reduced load, or "de-lid" the chip and use a better thermal compound. Obviously that last step voids the warranty, so only attempt that if you're very comfortable with the procedure and risks.

Linux 64bit i7-4770K CPU @ 3.50GHz host using the downloaded static version for the short tests. I will do the 2 SR5 tests but the remaining 2 will depend on challenge progress.

Starting Proth prime test of 2163*2^1255556+1 2163*2^1255556+1 is prime! (377964 decimal digits) Time : 387.274 sec.

Base factorized as : 2^2*17 Base prime factor(s) taken : 17 Starting N-1 prime test of 129897*68^129897+1 129897*68^129897+1 may be prime, trying to compute gcd's 3^((N-1)/17)-1 is coprime to N! 129897*68^129897+1 is prime! (238043 decimal digits) Time : 262.580 sec.

Starting Lucas Lehmer Riesel prime test of 1300274732577*2^1290000-1 Using zero-padded FMA3 FFT length 128K, Pass1=512, Pass2=256 1300274732577*2^1290000-1 is prime! (388341 decimal digits) Time : 540.096 sec.

Base prime factor(s) taken : 11 Starting N+1 prime test of 94112*121^94112-1 94112*121^94112-1 may be prime. Starting Lucas sequence... 94112*121^94112-1 may be prime, trying to compute gcd's Restarting Lucas sequence with P = 8 94112*121^94112-1 may be prime, trying to compute gcd's Restarting Lucas sequence with P = 9 94112*121^94112-1 may be prime, trying to compute gcd's U((N+1)/11) is coprime to N! 94112*121^94112-1 is prime! (196021 decimal digits, P = 9) Time : 1827.015 sec.

All of the Linux FMA3 manual prime tests are already done (see this Google Drive spreadsheet), but nobody is doing the BOINC tests for Linux FMA3 yet.
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74484 |

Prime Al Screen

Send message
Joined: 1 Dec 09
Posts: 240
ID: 50942
Credit: 38,208,946
RAC: 0

All of the Linux FMA3 manual prime tests are already done (see this Google Drive spreadsheet), but nobody is doing the BOINC tests for Linux FMA3 yet.

Not what I am seeing now or before.
Perhaps you need to update the spreadsheet so everyone can see what you see.

Toms, Anand and other sites have explained the heat issue that you describe. It was an observation of llr versions on the same host at the same time measured by the same tool so it not relevant to the cpu heatsink or measurement of said temperature.

ID: 74489 |

Usucapio Libertatis

Send message
Joined: 21 Apr 10
Posts: 746
ID: 59072
Credit: 702,637,573
RAC: 2,305

Mac Intel Core 2 Duo @ 2.16 GHz (Processor T7400)

Running 64 bit app:

$ ./llr64 -d -q"2163*2^1255556+1" Starting Proth prime test of 2163*2^1255556+1 Using all-complex FFT length 96K, Pass1=128, Pass2=768, a = 13 2163*2^1255556+1 is prime! (377964 decimal digits) Time : 2134.427 sec

.

Running 32 bit app:

./llr -d -q"2163*2^1255556+1" Starting Proth prime test of 2163*2^1255556+1 Using all-complex FFT length 96K, Pass1=128, Pass2=768, a = 13 2163*2^1255556+1 is prime! (377964 decimal digits) Time : 2203.456 sec.

____________
676754^262144+1 is prime

ID: 74495 |

JimB

Honorary cruncher
Send message
Joined: 4 Aug 11
Posts: 918
ID: 107307
Credit: 977,945,376
RAC: 45

All of the Linux FMA3 manual prime tests are already done (see this Google Drive spreadsheet), but nobody is doing the BOINC tests for Linux FMA3 yet.

Not what I am seeing now or before.
Perhaps you need to update the spreadsheet so everyone can see what you see.

It says "Artist" in every cell under the "Prime" column. That's the username of the person doing the tests. It's the BOINC column (labeled Completed and it's associated Validated column) that are still blank. It's been like that since yesterday.

ID: 74498 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

All of the Linux FMA3 manual prime tests are already done (see this Google Drive spreadsheet), but nobody is doing the BOINC tests for Linux FMA3 yet.

Not what I am seeing now or before.
Perhaps you need to update the spreadsheet so everyone can see what you see.

That's strange, because other people have said that not only can they see the spreadsheet, but they can actually see me moving the cursor from cell to cell as I'm making changes. There's no "saving"; the document gets updated in real time and everyone viewing the the document sees those changes in real time. I don't know why it's not working for you.

As of this moment, for Linux FMA3, all the manual prime tests are done (they're green with "Artist" in them), and all of the BOINC tests are available (the completed and validated columns are both red, except for GCW which is yellow meaning there's no suitable tests).

____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74499 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

Mac Intel Core 2 Duo @ 2.16 GHz (Processor T7400)

Running 64 bit app:

$ ./llr64 -d -q"2163*2^1255556+1" Starting Proth prime test of 2163*2^1255556+1 Using all-complex FFT length 96K, Pass1=128, Pass2=768, a = 13 2163*2^1255556+1 is prime! (377964 decimal digits) Time : 2134.427 sec
.

Running 32 bit app:

./llr -d -q"2163*2^1255556+1" Starting Proth prime test of 2163*2^1255556+1 Using all-complex FFT length 96K, Pass1=128, Pass2=768, a = 13 2163*2^1255556+1 is prime! (377964 decimal digits) Time : 2203.456 sec.

Is the spreadsheet not working? Is it not clear? I don't want people doing duplicate work unnecessarily.

The PPSE tests for both 32 and 64 bits on Mac are already done. (Gary for the 64 bit SSE3, and Iain for the 32 bit). As of this moment, only the SoB 32 bit manual prime test is available for 32 bit Mac, and for 64 bit SSE3 Mac only Sob, WOO, and both SR5 prime tests are available. On the BOINC tests, Iain is doing all the 32 bit tests, but all of the 64 bit SSE3 BOINC tests are available.

____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74500 |

Volunteer tester
Send message
Joined: 29 Sep 08
Posts: 87
ID: 29825
Credit: 365,054,285
RAC: 94,198

Is the spreadsheet not working? Is it not clear? I don't want people doing duplicate work unnecessarily.

The spreadsheet is fine. It is important to read http://www.primegrid.com/forum_thread.php?id=5557&nowrap=true#74345 to understand there are three parts of every test.

ID: 74502 |

Usucapio Libertatis

Send message
Joined: 21 Apr 10
Posts: 746
ID: 59072
Credit: 702,637,573
RAC: 2,305

I probably misundesrtood the color scheme on the spreadsheet. Looking at the columns, I also assumed you would be interested on having each test doublechecked. I' m pretty sure that the 32 bit test had no result or reserve when I started the test, a couple of hours before posting it here. Nevertheless, I apologise for any inconvenience I might have caused.
____________
676754^262144+1 is prime

ID: 74507 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

I probably misundesrtood the color scheme on the spreadsheet. Looking at the columns, I also assumed you would be interested on having each test doublechecked. I' m pretty sure that the 32 bit test had no result or reserve when I started the test, a couple of hours before posting it here. Nevertheless, I apologise for any inconvenience I might have caused.

That's why it's a good idea to tell me when you're starting the test rather as well as when you're completing the test. That's especially true for the longer tests which will take weeks on slower CPUs.

As for the colors, there is a legend on the spreadsheet that shows what each color means. I tried to make it as obvious as possible, but sometimes what I think is clear (as a native English speaker) isn't always as easiy to understand by people who speak English as a second language. (And, by the way, I'm totally amazed at how well many people here speak English! There's quite a few people whom I know as not native English speakers who are able to converse in written English so well it's almost impossible to tell it's not their native tongue.)

And, of course, sometimes I'm just as clear as mud. :)

Is there anything I can do to make the spreadsheet less confusing? Like, say, add instructions? (I think I'll go do that right now...)
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74509 |

Usucapio Libertatis

Send message
Joined: 21 Apr 10
Posts: 746
ID: 59072
Credit: 702,637,573
RAC: 2,305

Is there anything I can do to make the spreadsheet less confusing? Like, say, add instructions? (I think I'll go do that right now...)

English is obviously not my native tongue. However, looking at the spreadsheet now, on a 21 inch display, it looks clear as pure water. Earlier today, on my 13 MacBook, though, I totally missed both the legend on the top and the rest of the instructions (lines 7 to 9). I don't think there's any need to make it "less confusing".

I don't have any host matching the red tests, so I'll keep out of this thread.
____________
676754^262144+1 is prime

ID: 74510 |

Stojag

Send message
Joined: 8 Jun 10
Posts: 35
ID: 62023
Credit: 42,138,433
RAC: 0

Here's the last Prime test result for FMA3 Win64

C:\Users\Christof\Downloads>cllr64 -d -q"178658*5^1525224-1" Base prime factor(s) taken : 5 Starting N+1 prime test of 178658*5^1525224-1 Using FMA3 FFT length 384K, Pass1=256, Pass2=1536, a = 3 178658*5^1525224-1 may be prime. Starting Lucas sequence... Using FMA3 FFT length 384K, Pass1=256, Pass2=1536, P = 5 178658*5^1525224-1 may be prime, trying to compute gcd's Restarting Lucas sequence with P = 11 Using FMA3 FFT length 384K, Pass1=256, Pass2=1536, P = 11 178658*5^1525224-1 may be prime, trying to compute gcd's U((N+1)/5) is coprime to N! 178658*5^1525224-1 is prime! (1066092 decimal digits, P = 11) Time : 52555.902 sec.

ID: 74512 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

Here's the last Prime test result for FMA3 Win64

C:\Users\Christof\Downloads>cllr64 -d -q"178658*5^1525224-1" Base prime factor(s) taken : 5 Starting N+1 prime test of 178658*5^1525224-1 Using FMA3 FFT length 384K, Pass1=256, Pass2=1536, a = 3 178658*5^1525224-1 may be prime. Starting Lucas sequence... Using FMA3 FFT length 384K, Pass1=256, Pass2=1536, P = 5 178658*5^1525224-1 may be prime, trying to compute gcd's Restarting Lucas sequence with P = 11 Using FMA3 FFT length 384K, Pass1=256, Pass2=1536, P = 11 178658*5^1525224-1 may be prime, trying to compute gcd's U((N+1)/5) is coprime to N! 178658*5^1525224-1 is prime! (1066092 decimal digits, P = 11) Time : 52555.902 sec.

Thanks!
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74517 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

I've moved all the posts about FMA3 posting to this thread.
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74518 |

Scott Brown

Volunteer moderator
Project administrator
Volunteer tester
Project scientist
Avatar

Send message
Joined: 17 Oct 05
Posts: 2329
ID: 1178
Credit: 15,582,675,723
RAC: 15,114,994

32-bit SR5 +1 (Pentium M 2.1Ghz on Win7 Pro):

C:\Users\sbrow\Downloads\cllr38d>cllr.exe -d -q"59912*5^1500861+1" Base prime factor(s) taken : 5 Starting N-1 prime test of 59912*5^1500861+1 Using all-complex FFT length 384K, Pass1=256, Pass2=1536, a = 3 59912*5^1500861+1 may be prime, trying to compute gcd's 3^((N-1)/5)-1 is coprime to N! 59912*5^1500861+1 is prime! (1049062 decimal digits) Time : 89724.857 sec.

ID: 74521 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

32-bit SR5 +1 (Pentium M 2.1Ghz on Win7 Pro):

C:\Users\sbrow\Downloads\cllr38d>cllr.exe -d -q"59912*5^1500861+1" Base prime factor(s) taken : 5 Starting N-1 prime test of 59912*5^1500861+1 Using all-complex FFT length 384K, Pass1=256, Pass2=1536, a = 3 59912*5^1500861+1 may be prime, trying to compute gcd's 3^((N-1)/5)-1 is coprime to N! 59912*5^1500861+1 is prime! (1049062 decimal digits) Time : 89724.857 sec.

Thank you!

____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74524 |

pabliedung

Volunteer tester
Send message
Joined: 2 Jan 11
Posts: 87
ID: 79535
Credit: 262,308,907
RAC: 1

For anyone wishing to do the prime tests (i.e., running the new LLR against a known prime) I've added a list of known primes for each test to the spreadsheet. The list is over on the right side and you may need to scroll the spreadsheet to see it.

I have started the tests for Mac/AVX.

I will post the output from each test here as they complete.

ID: 74533 |

Volunteer tester
Send message
Joined: 29 Sep 08
Posts: 87
ID: 29825
Credit: 365,054,285
RAC: 94,198

but nobody is doing the BOINC tests for Linux FMA3 yet.

I would like to do it but i failed get it running. Maybe JimB can send me his configuration file and i try it one more time.

ID: 74535 |

JimB

Honorary cruncher
Send message
Joined: 4 Aug 11
Posts: 918
ID: 107307
Credit: 977,945,376
RAC: 45

I would like to do it but i failed get it running. Maybe JimB can send me his configuration file and i try it one more time.

PM sent.

ID: 74538 |

pabliedung

Volunteer tester
Send message
Joined: 2 Jan 11
Posts: 87
ID: 79535
Credit: 262,308,907
RAC: 1

For anyone wishing to do the prime tests (i.e., running the new LLR against a known prime) I've added a list of known primes for each test to the spreadsheet. The list is over on the right side and you may need to scroll the spreadsheet to see it.

I have started the tests for Mac/AVX.

I will post the output from each test here as they complete.

Here are the results so far. SoB and SR5 are still running.

iMaci7:LLR 3.8.13 test phil$ ./llr64 -d -q"2163*2^1255556+1" Starting Proth prime test of 2163*2^1255556+1 2163*2^1255556+1 is prime! (377964 decimal digits) Time : 820.572 sec. iMaci7:LLR 3.8.13 test phil$ ./llr64 -d -q"129897*68^129897+1" Base factorized as : 2^2*17 Base prime factor(s) taken : 17 Starting N-1 prime test of 129897*68^129897+1 129897*68^129897+1 may be prime, trying to compute gcd's 3^((N-1)/17)-1 is coprime to N! 129897*68^129897+1 is prime! (238043 decimal digits) Time : 393.985 sec. iMaci7:LLR 3.8.13 test phil$ ./llr64 -d -q"59912*5^1500861+1" Base prime factor(s) taken : 5 Starting N-1 prime test of 59912*5^1500861+1 59912*5^1500861+1 may be prime, trying to compute gcd's 3^((N-1)/5)-1 is coprime to N! 59912*5^1500861+1 is prime! (1049062 decimal digits) Time : 9047.166 sec. iMaci7:LLR 3.8.13 test phil$ ./llr64 -d -q"94112*121^94112-1" Base prime factor(s) taken : 11 Starting N+1 prime test of 94112*121^94112-1 94112*121^94112-1 may be prime. Starting Lucas sequence... 94112*121^94112-1 may be prime, trying to compute gcd's Restarting Lucas sequence with P = 8 94112*121^94112-1 may be prime, trying to compute gcd's Restarting Lucas sequence with P = 9 94112*121^94112-1 may be prime, trying to compute gcd's U((N+1)/11) is coprime to N! 94112*121^94112-1 is prime! (196021 decimal digits, P = 9) Time : 3271.203 sec. iMaci7:LLR 3.8.13 test phil$ ./llr64 -d -q"1300274732577*2^1290000-1" Starting Lucas Lehmer Riesel prime test of 1300274732577*2^1290000-1 Using zero-padded AVX FFT length 128K, Pass1=512, Pass2=256 1300274732577*2^1290000-1 is prime! (388341 decimal digits) Time : 948.414 sec. iMaci7:LLR 3.8.13 test phil$ ./llr64 -d -q"3752948*2^3752948-1" Starting Lucas Lehmer Riesel prime test of 3752948*2^3752948-1 Using zero-padded AVX FFT length 384K, Pass1=384, Pass2=1K 3752948*2^3752948-1 is prime! (1129757 decimal digits) Time : 9187.812 sec.

ID: 74541 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

For anyone wishing to do the prime tests (i.e., running the new LLR against a known prime) I've added a list of known primes for each test to the spreadsheet. The list is over on the right side and you may need to scroll the spreadsheet to see it.

I have started the tests for Mac/AVX.

I will post the output from each test here as they complete.

Here are the results so far. SoB and SR5 are still running.

Thanks!
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74542 |

pabliedung

Volunteer tester
Send message
Joined: 2 Jan 11
Posts: 87
ID: 79535
Credit: 262,308,907
RAC: 1

For anyone wishing to do the prime tests (i.e., running the new LLR against a known prime) I've added a list of known primes for each test to the spreadsheet. The list is over on the right side and you may need to scroll the spreadsheet to see it.

I have started the tests for Mac/AVX.

I will post the output from each test here as they complete.

Here are the results so far. SoB and SR5 are still running.

Thanks!

btw, all of the tests will be from this machine:

http://www.primegrid.com/show_host_detail.php?hostid=178104

ID: 74544 |

Honza

Volunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 15 Aug 05
Posts: 1931
ID: 352
Credit: 5,702,769,683
RAC: 1,041,236

manual AVX Windows, last one.
BOINC SoB should be finished tomorrow.

c:\_Honza>cllr64.exe -q"19249*2^13018586+1" -d Starting Proth prime test of 19249*2^13018586+1 Using all-complex AVX FFT length 1152K, Pass1=768, Pass2=1536, a = 3 19249*2^13018586+1 is prime! (3918990 decimal digits) Time : 158581.376 sec.

____________
My stats
Badge score: 1*1 + 5*1 + 8*3 + 9*11 + 10*1 + 11*1 + 12*3 = 186

ID: 74545 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

manual AVX Windows, last one.
BOINC SoB should be finished tomorrow.

c:\_Honza>cllr64.exe -q"19249*2^13018586+1" -d Starting Proth prime test of 19249*2^13018586+1 Using all-complex AVX FFT length 1152K, Pass1=768, Pass2=1536, a = 3 19249*2^13018586+1 is prime! (3918990 decimal digits) Time : 158581.376 sec.

Thanks!
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74551 |

Volunteer tester
Send message
Joined: 29 Sep 08
Posts: 87
ID: 29825
Credit: 365,054,285
RAC: 94,198

Thank you JimB, now it works!

386020184 PPSE (LLR)
386022477 Sophie Germain (LLR)
386020303 Sierpinski/Riesel Base 5 Problem (LLR) M
386023166 Sierpinski/Riesel Base 5 Problem (LLR) P
379967829 Seventeen or Bust
382019774 Woodall (LLR)

ID: 74556 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

Thank you JimB, now it works!

386020184 PPSE (LLR)
386022477 Sophie Germain (LLR)
386020303 Sierpinski/Riesel Base 5 Problem (LLR) M
386023166 Sierpinski/Riesel Base 5 Problem (LLR) P
379967829 Seventeen or Bust
382019774 Woodall (LLR)

Thanks!
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74557 |

Scott Brown

Volunteer moderator
Project administrator
Volunteer tester
Project scientist
Avatar

Send message
Joined: 17 Oct 05
Posts: 2329
ID: 1178
Credit: 15,582,675,723
RAC: 15,114,994

Got the 32-bit Linux laptop up and running (an older and even slower Pentium M than the Windows one I have doing the other tests), so I'll go ahead and start the small manual tests (i.e., PPSE, SGS, and both GCW tests). If all goes well over the next day or two (I need to see if this thing will still run stably), I'll go ahead with the SR5s as well.

ID: 74565 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

Got the 32-bit Linux laptop up and running (an older and even slower Pentium M than the Windows one I have doing the other tests), so I'll go ahead and start the small manual tests (i.e., PPSE, SGS, and both GCW tests). If all goes well over the next day or two (I need to see if this thing will still run stably), I'll go ahead with the SR5s as well.

I'll put you down for PPSE, SGS, GCW (+/-) and SR5 (+/-). The SR5 (-) test is pretty long due to the multiple passes needed to prove primality. It should take about 1/3rd the time of the SoB prime and five times as long as the Woodall prime. The credits listed at the bottom are indicative of the run times for those primes.

____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74569 |

Yankton
Send message
Joined: 16 Sep 08
Posts: 17
ID: 29071
Credit: 107,459,447
RAC: 6,350

Does anyone have a good app_info/app_config for using this app? Looks like I'm too late for testing it seperately but I can use it for cleanup work.

ID: 74585 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

Does anyone have a good app_info/app_config for using this app? Looks like I'm too late for testing it seperately but I can use it for cleanup work.

If you have a Haswell CPU you can get a decent speed boost. There's a moderate speed boost with Sandy Bridge and Ivy Bridge as well, and we always can use more reports of user experiences. Even if you can't do any of the organized testing, there's still plenty of reason to try out the new LLR app.

Here's the app_info I'm using, which has entries for all the LLR projects as well as both GeneferCUDA apps. (I don't have any of the sieves or the CPU or OpenCL apps in this file.) It's for 64 bit Windows, but on Linux or Mac or 32 bit Windows you just need to change the executable names.

<app_info> <app> <name>genefer</name> <user_friendly_name>Genefer</user_friendly_name> </app> <app> <name>genefer_wr</name> <user_friendly_name>Genefer (World Record)</user_friendly_name> </app> <app> <name>llrTPS</name> <user_friendly_name>Sophie Germain (LLR AVX2)</user_friendly_name> </app> <app> <name>llrWOO</name> <user_friendly_name>Woodall (LLR AVX2)</user_friendly_name> </app> <app> <name>llrCUL</name> <user_friendly_name>Cullen (LLR AVX2)</user_friendly_name> </app> <app> <name>llr321</name> <user_friendly_name>321 (LLR AVX2)</user_friendly_name> </app> <app> <name>llrPSP</name> <user_friendly_name>Prime Sierpinski Problem (LLR AVX2)</user_friendly_name> </app> <app> <name>llrPPS</name> <user_friendly_name>PPS (LLR AVX2)</user_friendly_name> </app> <app> <name>llrSOB</name> <user_friendly_name>Seventeen or Bust (LLR AVX2)</user_friendly_name> </app> <app> <name>llrTRP</name> <user_friendly_name>The Riesel Problem (LLR AVX2)</user_friendly_name> </app> <app> <name>llrPPSE</name> <user_friendly_name>PPSE (LLR AVX2)</user_friendly_name> </app> <app> <name>llrSR5</name> <user_friendly_name>Sierpinski/Riesel Base 5 Problem (LLR AVX)</user_friendly_name> </app> <file_info> <name>primegrid_genefer_3_1_2_9_2.12_windows_intelx86__cudaGFN.exe</name> <executable/> </file_info> <file_info> <name>cudart32_32_16.dll</name> </file_info> <file_info> <name>cufft32_32_16.dll</name> </file_info> <file_info> <name>primegrid_llr_wrapper.exe</name> <executable/> </file_info> <file_info> <name>cllr64.exe</name> <executable/> </file_info> <file_info> <name>llr.ini</name> </file_info> <app_version> <app_name>genefer</app_name> <version_num>008</version_num> <api_version>6.10.25</api_version> <file_ref> <file_name>primegrid_genefer_3_1_2_9_2.12_windows_intelx86__cudaGFN.exe</file_name> <main_program/> </file_ref> <file_ref> <file_name>cudart32_32_16.dll</file_name> <open_name>cudart32_32_16.dll</open_name> <copy_file/> </file_ref> <file_ref> <file_name>cufft32_32_16.dll</file_name> <open_name>cufft32_32_16.dll</open_name> <copy_file/> </file_ref> <plan_class>cudaGFN</plan_class> <coproc> <type>CUDA</type> <count>1.000000</count> </coproc> <avg_ncpus>0.1</avg_ncpus> <max_ncpus>0.1</max_ncpus> <gpu_ram>536870912.000000</gpu_ram> </app_version> <app_version> <app_name>genefer_wr</app_name> <version_num>007</version_num> <api_version>6.10.25</api_version> <file_ref> <file_name>primegrid_genefer_3_1_2_9_2.12_windows_intelx86__cudaGFN.exe</file_name> <main_program/> </file_ref> <file_ref> <file_name>cudart32_32_16.dll</file_name> <open_name>cudart32_32_16.dll</open_name> <copy_file/> </file_ref> <file_ref> <file_name>cufft32_32_16.dll</file_name> <open_name>cufft32_32_16.dll</open_name> <copy_file/> </file_ref> <plan_class>cudaGFNWR</plan_class> <coproc> <type>CUDA</type> <count>1.000000</count> </coproc> <avg_ncpus>0.1</avg_ncpus> <max_ncpus>0.1</max_ncpus> <gpu_ram>536870912.000000</gpu_ram> </app_version> <app_version> <app_name>llrTPS</app_name> <version_num>387</version_num> <api_version>6.10.6</api_version> <file_ref> <file_name>primegrid_llr_wrapper.exe</file_name> <main_program/> </file_ref> <file_ref> <file_name>cllr64.exe</file_name> <open_name>primegrid_cllr.exe.orig</open_name> </file_ref> <file_ref> <file_name>llr.ini</file_name> <open_name>llr.ini.6.00</open_name> </file_ref> </app_version> <app_version> <app_name>llrWOO</app_name> <version_num>387</version_num> <api_version>6.10.6</api_version> <file_ref> <file_name>primegrid_llr_wrapper.exe</file_name> <main_program/> </file_ref> <file_ref> <file_name>cllr64.exe</file_name> <open_name>primegrid_cllr.exe.orig</open_name> </file_ref> <file_ref> <file_name>llr.ini</file_name> <open_name>llr.ini.6.00</open_name> </file_ref> </app_version> <app_version> <app_name>llrCUL</app_name> <version_num>387</version_num> <api_version>6.10.6</api_version> <file_ref> <file_name>primegrid_llr_wrapper.exe</file_name> <main_program/> </file_ref> <file_ref> <file_name>cllr64.exe</file_name> <open_name>primegrid_cllr.exe.orig</open_name> </file_ref> <file_ref> <file_name>llr.ini</file_name> <open_name>llr.ini.6.00</open_name> </file_ref> </app_version> <app_version> <app_name>llr321</app_name> <version_num>387</version_num> <api_version>6.10.6</api_version> <file_ref> <file_name>primegrid_llr_wrapper.exe</file_name> <main_program/> </file_ref> <file_ref> <file_name>cllr64.exe</file_name> <open_name>primegrid_cllr.exe.orig</open_name> </file_ref> <file_ref> <file_name>llr.ini</file_name> <open_name>llr.ini.6.00</open_name> </file_ref> </app_version> <app_version> <app_name>llrPSP</app_name> <version_num>387</version_num> <api_version>6.10.6</api_version> <file_ref> <file_name>primegrid_llr_wrapper.exe</file_name> <main_program/> </file_ref> <file_ref> <file_name>cllr64.exe</file_name> <open_name>primegrid_cllr.exe.orig</open_name> </file_ref> <file_ref> <file_name>llr.ini</file_name> <open_name>llr.ini.6.00</open_name> </file_ref> </app_version> <app_version> <app_name>llrPPS</app_name> <version_num>387</version_num> <api_version>6.10.6</api_version> <file_ref> <file_name>primegrid_llr_wrapper.exe</file_name> <main_program/> </file_ref> <file_ref> <file_name>cllr64.exe</file_name> <open_name>primegrid_cllr.exe.orig</open_name> </file_ref> <file_ref> <file_name>llr.ini</file_name> <open_name>llr.ini.6.00</open_name> </file_ref> </app_version> <app_version> <app_name>llrSOB</app_name> <version_num>387</version_num> <api_version>6.10.6</api_version> <file_ref> <file_name>primegrid_llr_wrapper.exe</file_name> <main_program/> </file_ref> <file_ref> <file_name>cllr64.exe</file_name> <open_name>primegrid_cllr.exe.orig</open_name> </file_ref> <file_ref> <file_name>llr.ini</file_name> <open_name>llr.ini.6.00</open_name> </file_ref> </app_version> <app_version> <app_name>llrTRP</app_name> <version_num>387</version_num> <api_version>6.10.6</api_version> <file_ref> <file_name>primegrid_llr_wrapper.exe</file_name> <main_program/> </file_ref> <file_ref> <file_name>cllr64.exe</file_name> <open_name>primegrid_cllr.exe.orig</open_name> </file_ref> <file_ref> <file_name>llr.ini</file_name> <open_name>llr.ini.6.00</open_name> </file_ref> </app_version> <app_version> <app_name>llrPPSE</app_name> <version_num>387</version_num> <api_version>6.10.6</api_version> <file_ref> <file_name>primegrid_llr_wrapper.exe</file_name> <main_program/> </file_ref> <file_ref> <file_name>cllr64.exe</file_name> <open_name>primegrid_cllr.exe.orig</open_name> </file_ref> <file_ref> <file_name>llr.ini</file_name> <open_name>llr.ini.6.00</open_name> </file_ref> </app_version> <app_version> <app_name>llrSR5</app_name> <version_num>387</version_num> <api_version>6.10.6</api_version> <file_ref> <file_name>primegrid_llr_wrapper.exe</file_name> <main_program/> </file_ref> <file_ref> <file_name>cllr64.exe</file_name> <open_name>primegrid_cllr.exe.orig</open_name> </file_ref> <file_ref> <file_name>llr.ini</file_name> <open_name>llr.ini.6.00</open_name> </file_ref> </app_version> </app_info>

____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74587 |

Yankton
Send message
Joined: 16 Sep 08
Posts: 17
ID: 29071
Credit: 107,459,447
RAC: 6,350

I'd love to participate in the organized testing, don't get me wrong. Based on your spreadsheet though the contributions I could make have already been knocked out. I have2 haswell machines. One windows 7 (x64) and one linux (x64). If I can be of service there just let me know. I set primegrid for no new tasks. As soon as it runs out I'll fire this up and see how it goes!

ID: 74588 |

Scott Brown

Volunteer moderator
Project administrator
Volunteer tester
Project scientist
Avatar

Send message
Joined: 17 Oct 05
Posts: 2329
ID: 1178
Credit: 15,582,675,723
RAC: 15,114,994

32-bit PPSE Test (Pentium M on Linux)

sbrow@Jamie-Gateway-M320-and-4500-Series:~/Downloads$ ./sllr -d -q"2163*2^1255556+1" Starting Proth prime test of 2163*2^1255556+1 2163*2^1255556+1 is prime! (377964 decimal digits) Time : 8949.052 sec.

ID: 74590 |

Project administrator
Volunteer tester
Send message
Joined: 4 Dec 12
Posts: 1077
ID: 183129
Credit: 1,365,637,185
RAC: 1

I'll do PPSE, SGS, and SR5 -1 and +1 for mac OS X 64 bit.

that is, of course, if i can figure out how to use app_info.xml on mac.

help please?

thanks

____________

275*2^3585539+1 is prime!!! (1079358 digits)
Proud member of Aggie the Pew

ID: 74593 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

I'll do PPSE, SGS, and SR5 -1 and +1 for mac OS X 64 bit.

that is, of course, if i can figure out how to use app_info.xml on mac.

help please?

thanks

Put the app_info file into your BOINC/projects/www.primegrid.com directory. You can use the one I posted above, but you can remove the genefer stuff and change the names of the two executables. The llr executable you can download from Jean Penne's site (the link is in the other FMA thread.) The wrapper is the same as the stock wrapper.

Before activating app_info by creating the app_info.xml file, make sure you have no PrimeGrid tasks as they will be aborted.

You'll want to make a copy of the llr.ini.6.00 file and call it llr.ini.

____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74595 |

Project administrator
Volunteer tester
Send message
Joined: 4 Dec 12
Posts: 1077
ID: 183129
Credit: 1,365,637,185
RAC: 1

I'll do PPSE, SGS, and SR5 -1 and +1 for mac OS X 64 bit.

that is, of course, if i can figure out how to use app_info.xml on mac.

help please?

thanks

Put the app_info file into your BOINC/projects/www.primegrid.com directory. You can use the one I posted above, but you can remove the genefer stuff and change the names of the two executables. The llr executable you can download from Jean Penne's site (the link is in the other FMA thread.) The wrapper is the same as the stock wrapper.

Before activating app_info by creating the app_info.xml file, make sure you have no PrimeGrid tasks as they will be aborted.

You'll want to make a copy of the llr.ini.6.00 file and call it llr.ini.

Can't seem to get it to work. do i put the llr executable in www.primegrid.com directory also?

____________

275*2^3585539+1 is prime!!! (1079358 digits)

Proud member of Aggie the Pew

ID: 74598 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

I'll do PPSE, SGS, and SR5 -1 and +1 for mac OS X 64 bit.

that is, of course, if i can figure out how to use app_info.xml on mac.

help please?

thanks

Put the app_info file into your BOINC/projects/www.primegrid.com directory. You can use the one I posted above, but you can remove the genefer stuff and change the names of the two executables. The llr executable you can download from Jean Penne's site (the link is in the other FMA thread.) The wrapper is the same as the stock wrapper.

Before activating app_info by creating the app_info.xml file, make sure you have no PrimeGrid tasks as they will be aborted.

You'll want to make a copy of the llr.ini.6.00 file and call it llr.ini.

Can't seem to get it to work. do i put the llr executable in www.primegrid.com directory also?

All the files referenced in app_info must be in the www.primegrid.com directory. You can look at the beginning of your BOINC log to see if any files are missing. The warning message for missing files is red and easy to spot.
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74601 |

Iain Bethune

Honorary cruncher
Send message
Joined: 28 Jan 09
Posts: 1588
ID: 34775
Credit: 194,056,043
RAC: 0

I'll do PPSE, SGS, and SR5 -1 and +1 for mac OS X 64 bit.

that is, of course, if i can figure out how to use app_info.xml on mac.

help please?

thanks

Here is an app_info.xml for Mac. It's slightly different to the windows one in terms of the filenames. If you have any more problems you can PM me or post here:

<app_info> <app> <name>llrTPS</name> <user_friendly_name>Sophie Germain Prime Search (LLR)</user_friendly_name> </app> <app> <name>llrPPSE</name> <user_friendly_name>PPSE (LLR)</user_friendly_name> </app> <app> <name>llrSR5</name> <user_friendly_name>SR5 (LLR)</user_friendly_name> </app> <app> <name>llrWOO</name> <user_friendly_name>WOO (LLR)</user_friendly_name> </app> <app> <name>llrSOB</name> <user_friendly_name>SOB (LLR)</user_friendly_name> </app> <file_info> <name>primegrid_llr_wrapper_6.16_x86_64-apple-darwin</name> <executable/> </file_info> <file_info> <name>llr64</name> <executable/> </file_info> <file_info> <name>llr.ini.6.00</name> </file_info> <app_version> <app_name>llrTPS</app_name> <version_num>616</version_num> <file_ref> <file_name>primegrid_llr_wrapper_6.16_x86_64-apple-darwin</file_name> <main_program/> </file_ref> <file_ref> <file_name>llr64</file_name> <open_name>primegrid_llr.orig</open_name> </file_ref> <file_ref> <file_name>llr.ini.6.00</file_name> <open_name>llr.ini.orig</open_name> </file_ref> </app_version> <app_version> <app_name>llrPPSE</app_name> <version_num>616</version_num> <file_ref> <file_name>primegrid_llr_wrapper_6.16_x86_64-apple-darwin</file_name> <main_program/> </file_ref> <file_ref> <file_name>llr64</file_name> <open_name>primegrid_llr.orig</open_name> </file_ref> <file_ref> <file_name>llr.ini.6.00</file_name> <open_name>llr.ini.orig</open_name> </file_ref> </app_version> <app_version> <app_name>llrSR5</app_name> <version_num>616</version_num> <file_ref> <file_name>primegrid_llr_wrapper_6.16_x86_64-apple-darwin</file_name> <main_program/> </file_ref> <file_ref> <file_name>llr64</file_name> <open_name>primegrid_llr.orig</open_name> </file_ref> <file_ref> <file_name>llr.ini.6.00</file_name> <open_name>llr.ini.orig</open_name> </file_ref> </app_version> <app_version> <app_name>llrWOO</app_name> <version_num>616</version_num> <file_ref> <file_name>primegrid_llr_wrapper_6.16_x86_64-apple-darwin</file_name> <main_program/> </file_ref> <file_ref> <file_name>llr64</file_name> <open_name>primegrid_llr.orig</open_name> </file_ref> <file_ref> <file_name>llr.ini.6.00</file_name> <open_name>llr.ini.orig</open_name> </file_ref> </app_version> <app_version> <app_name>llrSOB</app_name> <version_num>616</version_num> <file_ref> <file_name>primegrid_llr_wrapper_6.16_x86_64-apple-darwin</file_name> <main_program/> </file_ref> <file_ref> <file_name>llr64</file_name> <open_name>primegrid_llr.orig</open_name> </file_ref> <file_ref> <file_name>llr.ini.6.00</file_name> <open_name>llr.ini.orig</open_name> </file_ref> </app_version> </app_info>

Cheers

- Iain
____________
Twitter: IainBethune
Proud member of team "Aggie The Pew". Go Aggie!
3073428256125*2^1290000-1 is Prime!

ID: 74608 |

Scott Brown

Volunteer moderator
Project administrator
Volunteer tester
Project scientist
Avatar

Send message
Joined: 17 Oct 05
Posts: 2329
ID: 1178
Credit: 15,582,675,723
RAC: 15,114,994

32-bit GCW +1 Test (Pentium M on Linux)

sbrow@Jamie-Gateway-M320-and-4500-Series:~/Downloads$ ./sllr -d -q"129897*68^129897+1" Base factorized as : 2^2*17 Base prime factor(s) taken : 17 Starting N-1 prime test of 129897*68^129897+1 129897*68^129897+1 may be prime, trying to compute gcd's 3^((N-1)/17)-1 is coprime to N! 129897*68^129897+1 is prime! (238043 decimal digits) Time : 6946.732 sec.

ID: 74611 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

Thank you!
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74615 |

Honza

Volunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 15 Aug 05
Posts: 1931
ID: 352
Credit: 5,702,769,683
RAC: 1,041,236

Windows AVX BOINC SoB finished and validated.

http://www.primegrid.com/workunit.php?wuid=384279920

____________
My stats
Badge score: 1*1 + 5*1 + 8*3 + 9*11 + 10*1 + 11*1 + 12*3 = 186

ID: 74622 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

Windows AVX BOINC SoB finished and validated.

http://www.primegrid.com/workunit.php?wuid=384279920

Thanks!
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74623 |

pabliedung

Volunteer tester
Send message
Joined: 2 Jan 11
Posts: 87
ID: 79535
Credit: 262,308,907
RAC: 1

For anyone wishing to do the prime tests (i.e., running the new LLR against a known prime) I've added a list of known primes for each test to the spreadsheet. The list is over on the right side and you may need to scroll the spreadsheet to see it.

I have started the tests for Mac/AVX.

I will post the output from each test here as they complete.

Here are the results so far. SoB and SR5 are still running.

Thanks!

SoB and SR5 are finished now too:

iMaci7:LLR 3.8.13 test phil$ ./llr64 -d -q"19249*2^13018586+1" Starting Proth prime test of 19249*2^13018586+1 19249*2^13018586+1 is prime! (3918990 decimal digits) Time : 116297.853 sec. iMaci7:LLR 3.8.13 test phil$ ./llr64 -d -q"178658*5^1525224-1" Base prime factor(s) taken : 5 Starting N+1 prime test of 178658*5^1525224-1 178658*5^1525224-1 may be prime. Starting Lucas sequence... 178658*5^1525224-1 may be prime, trying to compute gcd's Restarting Lucas sequence with P = 11 178658*5^1525224-1 may be prime, trying to compute gcd's U((N+1)/5) is coprime to N! 178658*5^1525224-1 is prime! (1066092 decimal digits, P = 11) Time : 86288.802 sec.

ID: 74625 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

For anyone wishing to do the prime tests (i.e., running the new LLR against a known prime) I've added a list of known primes for each test to the spreadsheet. The list is over on the right side and you may need to scroll the spreadsheet to see it.

I have started the tests for Mac/AVX.

I will post the output from each test here as they complete.

Here are the results so far. SoB and SR5 are still running.

Thanks!

SoB and SR5 are finished now too:

iMaci7:LLR 3.8.13 test phil$ ./llr64 -d -q"19249*2^13018586+1" Starting Proth prime test of 19249*2^13018586+1 19249*2^13018586+1 is prime! (3918990 decimal digits) Time : 116297.853 sec. iMaci7:LLR 3.8.13 test phil$ ./llr64 -d -q"178658*5^1525224-1" Base prime factor(s) taken : 5 Starting N+1 prime test of 178658*5^1525224-1 178658*5^1525224-1 may be prime. Starting Lucas sequence... 178658*5^1525224-1 may be prime, trying to compute gcd's Restarting Lucas sequence with P = 11 178658*5^1525224-1 may be prime, trying to compute gcd's U((N+1)/5) is coprime to N! 178658*5^1525224-1 is prime! (1066092 decimal digits, P = 11) Time : 86288.802 sec.

Thanks!
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74629 |

pabliedung

Volunteer tester
Send message
Joined: 2 Jan 11
Posts: 87
ID: 79535
Credit: 262,308,907
RAC: 1

I'll do PPSE, SGS, and SR5 -1 and +1 for mac OS X 64 bit.

that is, of course, if i can figure out how to use app_info.xml on mac.

help please?

thanks

Here is an app_info.xml for Mac. It's slightly different to the windows one in terms of the filenames. If you have any more problems you can PM me or post here:

Thanks Iain for posting the code for Mac. I had some trouble earlier adapting the app_info file... I got it to work now.

I have started to double check a SoB and Woodall workunit with the 3.8.13 version on Mac/AVX.

SoB: http://www.primegrid.com/workunit.php?wuid=380204928
WOO: http://www.primegrid.com/workunit.php?wuid=383077325

ID: 74634 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

I'll do PPSE, SGS, and SR5 -1 and +1 for mac OS X 64 bit.

that is, of course, if i can figure out how to use app_info.xml on mac.

help please?

thanks

Here is an app_info.xml for Mac. It's slightly different to the windows one in terms of the filenames. If you have any more problems you can PM me or post here:

Thanks Iain for posting the code for Mac. I had some trouble earlier adapting the app_info file... I got it to work now.

I have started to double check a SoB and Woodall workunit with the 3.8.13 version on Mac/AVX.

SoB: http://www.primegrid.com/workunit.php?wuid=380204928
WOO: http://www.primegrid.com/workunit.php?wuid=383077325

Thanks
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74635 |

pabliedung

Volunteer tester
Send message
Joined: 2 Jan 11
Posts: 87
ID: 79535
Credit: 262,308,907
RAC: 1

below are validated wu's for PPSE and SGS with the 3.8.13 version on Mac/AVX.

PPSE: http://www.primegrid.com/workunit.php?wuid=386146603 (verified)
SGS: http://www.primegrid.com/workunit.php?wuid=385850182 (verified)

SR5's are currently pending ... I will try to get some wu's where I am the double checker so we have the results faster. I'll post them shortly.

SR5 (+1): http://www.primegrid.com/workunit.php?wuid=386153651 (pending)
SR5 (-1): http://www.primegrid.com/workunit.php?wuid=386153117 (pending)

ID: 74650 |

pabliedung

Volunteer tester
Send message
Joined: 2 Jan 11
Posts: 87
ID: 79535
Credit: 262,308,907
RAC: 1

Found two where I am the double checker:

SR5 (-1): http://www.primegrid.com/workunit.php?wuid=385821443
SR5 (+1): http://www.primegrid.com/workunit.php?wuid=386166926

Results should come in within the next 4h ... unless the two previously linked wu's validate first.

ID: 74652 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

Found two where I am the double checker:

SR5 (-1): http://www.primegrid.com/workunit.php?wuid=385821443
SR5 (+1): http://www.primegrid.com/workunit.php?wuid=386166926

Results should come in within the next 4h ... unless the two previously linked wu's validate first.

Ok, I'll use these.

As an added bonus, the +1 is one of those odd "double check" tasks where the first task is a residue from an external source, so we're only running a single task against the pre-existing result.

For the record, I recommended doing this for the SoB because those can sometimes take months, and possibly the Woodall, but for all the rest the wait won't be too long. Don't forget we're waiting on non-AVX SoB tests to complete, and that will probably take about 2 weeks. Also, this is NOT the last step in putting these apps live, so we won't get hurt by waiting a week or three for the tasks to validate.
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74653 |

pabliedung

Volunteer tester
Send message
Joined: 2 Jan 11
Posts: 87
ID: 79535
Credit: 262,308,907
RAC: 1

For the record, I recommended doing this for the SoB because those can sometimes take months, and possibly the Woodall, but for all the rest the wait won't be too long. Don't forget we're waiting on non-AVX SoB tests to complete, and that will probably take about 2 weeks. Also, this is NOT the last step in putting these apps live, so we won't get hurt by waiting a week or three for the tasks to validate.

I am already the double checker for both the SoB and Woodall wu's ... those should hopefully validate right away.

ID: 74654 |

Scott Brown

Volunteer moderator
Project administrator
Volunteer tester
Project scientist
Avatar

Send message
Joined: 17 Oct 05
Posts: 2329
ID: 1178
Credit: 15,582,675,723
RAC: 15,114,994

32-bit SGS Test (Pentium M on Linux)

sbrow@Jamie-Gateway-M320-and-4500-Series:~/Downloads$ ./sllr -d -q"1300274732577*2^1290000-1" Starting Lucas Lehmer Riesel prime test of 1300274732577*2^1290000-1 Using zero-padded FFT length 128K, Pass1=512, Pass2=256 1300274732577*2^1290000-1 is prime! (388341 decimal digits) Time : 12078.635 sec.

ID: 74655 |

Volunteer moderator
Project administrator

Send message
Joined: 21 Jan 10
Posts: 13804
ID: 53948
Credit: 345,369,032
RAC: 4,797

32-bit SGS Test (Pentium M on Linux)

sbrow@Jamie-Gateway-M320-and-4500-Series:~/Downloads$ ./sllr -d -q"1300274732577*2^1290000-1" Starting Lucas Lehmer Riesel prime test of 1300274732577*2^1290000-1 Using zero-padded FFT length 128K, Pass1=512, Pass2=256 1300274732577*2^1290000-1 is prime! (388341 decimal digits) Time : 12078.635 sec.

Thank you!
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

ID: 74656 |

Project administrator
Volunteer tester
Send message
Joined: 4 Dec 12
Posts: 1077
ID: 183129
Credit: 1,365,637,185
RAC: 1

I'll do PPSE, SGS, and SR5 -1 and +1 for mac OS X 64 bit.

that is, of course, if i can figure out how to use app_info.xml on mac.

help please?

thanks

Here is an app_info.xml for Mac. It's slightly different to the windows one in terms of the filenames. If you have any more problems you can PM me or post here:

<app_info> <app> <name>llrTPS</name> <user_friendly_name>Sophie Germain Prime Search (LLR)</user_friendly_name> </app> <app> <name>llrPPSE</name> <user_friendly_name>PPSE (LLR)</user_friendly_name> </app> <app> <name>llrSR5</name> <user_friendly_name>SR5 (LLR)</user_friendly_name> </app> <app> <name>llrWOO</name> <user_friendly_name>WOO (LLR)</user_friendly_name> </app> <app> <name>llrSOB</name> <user_friendly_name>SOB (LLR)</user_friendly_name> </app> <file_info> <name>primegrid_llr_wrapper_6.16_x86_64-apple-darwin</name> <executable/> </file_info> <file_info> <name>llr64</name> <executable/> </file_info> <file_info> <name>llr.ini.6.00</name> </file_info> <app_version> <app_name>llrTPS</app_name> <version_num>616</version_num> <file_ref> <file_name>primegrid_llr_wrapper_6.16_x86_64-apple-darwin</file_name> <main_program/> </file_ref> <file_ref> <file_name>llr64</file_name> <open_name>primegrid_llr.orig</open_name> </file_ref> <file_ref> <file_name>llr.ini.6.00</file_name> <open_name>llr.ini.orig</open_name> </file_ref> </app_version> <app_version> <app_name>llrPPSE</app_name> <version_num>616</version_num> <file_ref> <file_name>primegrid_llr_wrapper_6.16_x86_64-apple-darwin</file_name> <main_program/> </file_ref> <file_ref> <file_name>llr64</file_name> <open_name>primegrid_llr.orig</open_name> </file_ref> <file_ref> <file_name>llr.ini.6.00</file_name> <open_name>llr.ini.orig</open_name> </file_ref> </app_version> <app_version> <app_name>llrSR5</app_name> <version_num>616</version_num> <file_ref> <file_name>primegrid_llr_wrapper_6.16_x86_64-apple-darwin</file_name> <main_program/> </file_ref> <file_ref> <file_name>llr64</file_name> <open_name>primegrid_llr.orig</open_name> </file_ref> <file_ref> <file_name>llr.ini.6.00</file_name> <open_name>llr.ini.orig</open_name> </file_ref> </app_version> <app_version> <app_name>llrWOO</app_name> <version_num>616</version_num> <file_ref> <file_name>primegrid_llr_wrapper_6.16_x86_64-apple-darwin</file_name> <main_program/> </file_ref> <file_ref> <file_name>llr64</file_name> <open_name>primegrid_llr.orig</open_name> </file_ref> <file_ref> <file_name>llr.ini.6.00</file_name> <open_name>llr.ini.orig</open_name> </file_ref> </app_version> <app_version> <app_name>llrSOB</app_name> <version_num>616</version_num> <file_ref> <file_name>primegrid_llr_wrapper_6.16_x86_64-apple-darwin</file_name> <main_program/> </file_ref> <file_ref> <file_name>llr64</file_name> <open_name>primegrid_llr.orig</open_name> </file_ref> <file_ref> <file_name>llr.ini.6.00</file_name> <open_name>llr.ini.orig</open_name> </file_ref> </app_version> </app_info>

Cheers

- Iain

It is not working. (the app_info is, the llr version is not)
The primegrid llr version is working, i confirm that the FMA3 llr version is not working. It simply runs for a few seconds then stops.
System specs: i5 2500k, 8gb ram, Mac OS X 10.8.2 64 bit.

Here is an example of one of the WUs: http://www.primegrid.com/workunit.php?wuid=386171753
and the computer ID is 428747
Any help would be appreciated.
thanks,
-1998golfer
____________

275*2^3585539+1 is prime!!! (1079358 digits)

Proud member of Aggie the Pew

ID: 74658 |

pabliedung

Volunteer tester
Send message
Joined: 2 Jan 11
Posts: 87
ID: 79535
Credit: 262,308,907
RAC: 1

It is not working. (the app_info is, the llr version is not)
The primegrid llr version is working, i confirm that the FMA3 llr version is not working. It simply runs for a few seconds then stops.
System specs: i5 2500k, 8gb ram, Mac OS X 10.8.2 64 bit.

Here is an example of one of the WUs: http://www.primegrid.com/workunit.php?wuid=386171753
and the computer ID is 428747
Any help would be appreciated.
thanks,
-1998golfer

The Stderr output suggests the llr file that you copied into the primegrid directory doesn't have the same rights as the other Boinc files and/or doesn't belong to the proper Boinc user group.

I had to make the following changes manually:

change access rights:

chmod 664 app_info.xml

chmod 664 llr64

and change user groups:

chown boinc_master:boinc_project app_info.xml

chown boinc_master:boinc_project llr64

run these commands as the super user ("sudo -i") in the primegrid folder and see if that fixes the setuid and setgid errors.

ID: 74659 |

Project administrator
Volunteer tester
Send message
Joined: 4 Dec 12
Posts: 1077
ID: 183129
Credit: 1,365,637,185
RAC: 1

It is not working. (the app_info is, the llr version is not)
The primegrid llr version is working, i confirm that the FMA3 llr version is not working. It simply runs for a few seconds then stops.
System specs: i5 2500k, 8gb ram, Mac OS X 10.8.2 64 bit.

Here is an example of one of the WUs: http://www.primegrid.com/workunit.php?wuid=386171753
and the computer ID is 428747
Any help would be appreciated.
thanks,
-1998golfer

The Stderr output suggests the llr file that you copied into the primegrid directory doesn't have the same rights as the other Boinc files and/or doesn't belong to the proper Boinc user group.

I had to make the following changes manually:

change access rights:

chmod 664 app_info.xml

chmod 664 llr64

and change user groups:

chown boinc_master:boinc_project app_info.xml

chown boinc_master:boinc_project llr64

run these commands as the super user ("sudo -i") in the primegrid folder and see if that fixes the setuid and setgid errors.

Thank you, I have to go for now but I'll try that tomorrow.

____________

275*2^3585539+1 is prime!!! (1079358 digits)

Proud member of Aggie the Pew

ID: 74660 |

Iain Bethune

Honorary cruncher
Send message
Joined: 28 Jan 09
Posts: 1588
ID: 34775
Credit: 194,056,043
RAC: 0

Hi 1998golfer,

The output does suggest the app is running OK, most likely it's failing because the llr.ini file is missing or named incorrectly. The files you need should be in the /projects/www.primegrid.com directory and named according to the <file_info> entries in the app_info.xml. You'll also need to set the permissions to match the other files.

Cheers

- Iain
____________
Twitter: IainBethune
Proud member of team "Aggie The Pew". Go Aggie!
3073428256125*2^1290000-1 is Prime!

ID: 74669 |

Gary Craig
Volunteer tester
Avatar

Send message
Joined: 30 Dec 09
Posts: 3213
ID: 52890
Credit: 1,005,618,748
RAC: 0

Signing up for the six BOINC tests on Mac-SSE3.

First two tests (the short ones, of course) already complete and uploaded (both still waiting for validation, at least as I type this):
PPSE: WU 386233830
SGS: WU 386237365

Have downloaded Woodall WU 382310381 and it's running. See you in 3 days! :-)

--Gary

ID: 74685 |