mersenneforum.org

mersenneforum.org (https://www.mersenneforum.org/index.php)
-   Software (https://www.mersenneforum.org/forumdisplay.php?f=10)
-   -   Other P-1 implementations? (https://www.mersenneforum.org/showthread.php?t=27637)

mathwiz 2022-03-07 14:50

Other P-1 implementations?
 
Are there other libraries that implement P-1 besides mprime that are reasonably fast (if not quite as fast) and also fairly easy to read/follow the code? Maybe even a library that can be called into.

Uncwilly 2022-03-07 15:54

gpuOwl and mlucas both do P-1

ixfd64 2022-03-07 16:52

GMP-ECM does too, but only for small numbers.

kriesel 2022-03-07 20:09

[QUOTE=mathwiz;601267]Are there other libraries that implement P-1 besides mprime that are reasonably fast (if not quite as fast) and also fairly easy to read/follow the code? Maybe even a library that can be called into.[/QUOTE]
You don't say what type or size of number you wish to factor. But assuming largish Mersenne numbers, the reference info has [URL="https://www.mersenneforum.org/showpost.php?p=488291&postcount=2"]this post[/URL] with an attachment which is periodically updated.
Corresponding to the mprime / prime95 applications, there is the gwnum library of routines.
Others have mentioned gpuowl and Mlucas and gmp_ecm applications, & there's CUDAPm1 & [URL="https://gmplib.org/"]GMP[/URL].

There are also some python or perl extended-precision implementations around, and CUDA and OpenCL fft library implementations from which P-1 applications can be developed, and tools like Pari/GP or Mathematica etc. There are a few examples of implementations known to exist, such as LaurV's: [URL]https://mersenneforum.org/showpost.php?p=308733&postcount=11[/URL] [URL]https://www.mersenneforum.org/showpost.php?p=341971&postcount=277[/URL]
IIRC they were developed more for learning than production.

tuckerkao 2022-04-24 09:43

[QUOTE=Uncwilly;601274]gpuOwl and mlucas both do P-1[/QUOTE]
Will they consume 99% of the GPU resource and only 1% of the CPU resource like mfaktc? Or will they slow down the Prime95 assignments by a considerable ratio?

I still don't know how to run P-1 on GpuOwl with the Nvidia GPU. What line should I type in to worktodo.txt?

Uncwilly 2022-04-24 13:56

[QUOTE=tuckerkao;604674]I still don't know how to run P-1 on GpuOwl with the Nvidia GPU. What line should I type in to worktodo.txt?[/QUOTE]Go find the answer for yourself here: [URL="https://www.mersenneforum.org/forumdisplay.php?f=154"]https://www.mersenneforum.org/forumdisplay.php?f=154 [/URL]

kriesel 2022-04-24 15:14

Yes, please use the existing provided resources. And [QUOTE=Uncwilly;604684][URL="https://www.mersenneforum.org/forumdisplay.php?f=154"]https://www.mersenneforum.org/forumdisplay.php?f=154 [/URL][/QUOTE]is a haystack;
[URL]https://mersenneforum.org/showthread.php?t=24607[/URL] is its table of contents (or two) with links to use. For those too impatient to read or skim a table of contents, there is a handy feature in most if not all modern browsers, to search for a keyword or phrase in the currently viewed web page. (In Firefox, as easy and fast as Ctrl-F)

tuckerkao 2022-04-24 17:23

[QUOTE=Uncwilly;604684]Go find the answer for yourself here:[/QUOTE]
The post I eventually found was [URL="https://www.mersenneforum.org/showpost.php?p=522098&postcount=22"]https://www.mersenneforum.org/showpost.php?p=522098&postcount=22[/URL]. It showed Pfactor instead of Pminus1, maybe this has been the reason why I always not be able to reserve the P-1 assignments of Prime95 from the server.


All times are UTC. The time now is 14:09.

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2023, Jelsoft Enterprises Ltd.