mersenneforum.org  

Go Back   mersenneforum.org > Great Internet Mersenne Prime Search > Hardware > GPU Computing

Reply
 
Thread Tools
Old 2013-11-03, 02:13   #1
fairsky
 
Sep 2013

41 Posts
Default how to get cudaLucas work for large exponent?

For example, I had a few jobs which are large exponents, such as 444xxxxxx from 78bit.

If I use fft length automatically from the option
Quote:
FFTLength=0
in setting file, the tesla cannot work but pop out the error
Quote:
over specifications Grid = 98304
Have you encountered this error before? How to get it work, any idea?
fairsky is offline   Reply With Quote
Old 2013-11-05, 20:41   #2
owftheevil
 
owftheevil's Avatar
 
"Carl Darby"
Oct 2012
Spring Mountains, Nevada

32·5·7 Posts
Default

Try using Threads=512 in the ini file.
owftheevil is offline   Reply With Quote
Old 2013-11-05, 21:57   #3
fairsky
 
Sep 2013

41 Posts
Default

Quote:
Originally Posted by owftheevil View Post
Try using Threads=512 in the ini file.
Thank you for the suggestion, but I tried even from 32 to 1024, for exponent 4xxxxxxxx.

And even for exponent 322xxxxxx with depth=76 or above I will also fail to run the cudaLucas ...
fairsky is offline   Reply With Quote
Old 2013-11-21, 17:27   #4
Robish
 
"Rob Gahan"
Aug 2013
Ireland

22·32 Posts
Smile

Quote:
Originally Posted by fairsky View Post
For example, I had a few jobs which are large exponents, such as 444xxxxxx from 78bit.

If I use fft length automatically from the option

in setting file, the tesla cannot work but pop out the error


Have you encountered this error before? How to get it work, any idea?
I use Cudalucas.x.x.x.x -threads 512 -f 20971520 -t 332xxxxxx.

However my exponents are smaller than yours so -f may need to increase, 20971520 = 1024 * 20480. So multiples of 1024. I dont fully understand it, but I would try 1024 * 3072 = 31457280 if 20971520 doesnt work.

LaurV explained it but I never got a chance to study it yet.

http://www.mersenneforum.org/showpos...&postcount=236

Hope this helps

Cheers

Rob.
Robish is offline   Reply With Quote
Old 2013-11-21, 17:32   #5
Robish
 
"Rob Gahan"
Aug 2013
Ireland

22·32 Posts
Smile

Quote:
Originally Posted by Robish View Post
I use Cudalucas.x.x.x.x -threads 512 -f 20971520 -t 332xxxxxx.

However my exponents are smaller than yours so -f may need to increase, 20971520 = 1024 * 20480. So multiples of 1024. I dont fully understand it, but I would try 1024 * 3072 = 31457280 if 20971520 doesnt work.

LaurV explained it but I never got a chance to study it yet.

http://www.mersenneforum.org/showpos...&postcount=236

Hope this helps

Cheers

Rob.

Be warned, Im using GTX690's and 332xxxxxx takes 4500 hours or 190 days to run, hope your not in a hurry, :-P
Robish is offline   Reply With Quote
Old 2013-11-22, 10:33   #6
Manpowre
 
"Svein Johansen"
May 2013
Norway

3·67 Posts
Default

Quote:
Originally Posted by Robish View Post
Be warned, Im using GTX690's and 332xxxxxx takes 4500 hours or 190 days to run, hope your not in a hurry, :-P
really 4500 hours ? LOL..
Manpowre is offline   Reply With Quote
Reply

Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
CudaLucas Residual evoflash GPU Computing 21 2017-11-13 12:04
Don't DC/LL them with CudaLucas LaurV Data 131 2017-05-02 18:41
CUDALucas gives all-zero residues fivemack GPU Computing 4 2016-07-21 15:49
Automated work for cudaLucas bgbeuning GPU Computing 1 2016-02-20 01:34
Trying to run CUDALucas on Windows 8 CP Rodrigo GPU Computing 12 2012-03-07 23:20

All times are UTC. The time now is 15:09.


Tue Aug 3 15:09:02 UTC 2021 up 11 days, 9:38, 1 user, load averages: 3.69, 4.35, 4.30

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2021, Jelsoft Enterprises Ltd.

This forum has received and complied with 0 (zero) government requests for information.

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.
A copy of the license is included in the FAQ.