mersenneforum.org  

Go Back   mersenneforum.org > Search Forums

Showing results 1 to 25 of 1000
Search took 0.47 seconds.
Search: Posts Made By: jasonp
Forum: Factoring 2020-06-10, 11:36
Replies: 13
Views: 564
Posted By jasonp
That is how Msieve works. We have many...

That is how Msieve works. We have many experienced programmers here and many of them have also implemented some version of QS. Start a thread asking questions if you like, but also check the archives...
Forum: Factoring 2020-06-09, 18:43
Replies: 13
Views: 564
Posted By jasonp
You should be prepared to manually force a...

You should be prepared to manually force a remainder whose sign you want when the numerator of a modulo operation is negative, because a case can be made for both a positive and negative remainder....
Forum: CADO-NFS 2020-05-21, 16:59
Replies: 19
Views: 718
Posted By jasonp
The limit on admax is 1/(poly_degree+1) the size...

The limit on admax is 1/(poly_degree+1) the size of the number to be factored. That's the extreme upper limit; Kleinjung's 2006 paper gives more sensible bounds on admax based on the largest and...
Forum: YAFU 2020-05-21, 16:49
Replies: 35
Views: 5,329
Posted By jasonp
Percival predicted a factor of two speedup from...

Percival predicted a factor of two speedup from the method; maybe the factor is larger if one uses a much larger factor base to improve the relation yield and use the subset sum technique to avoid...
Forum: YAFU 2020-05-20, 17:00
Replies: 35
Views: 5,329
Posted By jasonp
Colin Percival, in a blog post around ~2005...

Colin Percival, in a blog post around ~2005 (https://www.daemonology.net/blog/2005-11-30-quadratic-sieve-constant.html), called this method the subset sum self initializing quadratic sieve (SSSIQS)....
Forum: Msieve 2020-05-15, 13:41
Replies: 3
Views: 177
Posted By jasonp
You will need a version of CUDA that supports...

You will need a version of CUDA that supports compute capability 7.5 and a version of Visual Studio that that version of CUDA is compatible with. If you have those, both the Msieve makefile and...
Forum: Lounge 2020-05-08, 16:00
Replies: 11
Views: 835
Posted By jasonp
Thank you, this is wonderful.

Thank you, this is wonderful.
Forum: Msieve 2020-04-15, 16:52
Replies: 11
Views: 854
Posted By jasonp
Alpertron has worked in this area for many years,...

Alpertron has worked in this area for many years, and I assume his work would be difficult to outrun. That being said a 300 bit number in around 20 minutes on my laptop is surprisingly good

I hope...
Forum: Factoring 2020-02-20, 14:31
Replies: 42
Views: 3,770
Posted By jasonp
Maybe Ilya Popvyan (poily on the forum here) was...

Maybe Ilya Popvyan (poily on the forum here) was involved. He contributed Lanczos optimizations several years ago and I know he was running jobs on this machine.
Forum: Factoring 2020-01-13, 04:10
Replies: 24
Views: 3,220
Posted By jasonp
I kind of wish I had been involved in the...

I kind of wish I had been involved in the factoring field back in 2000, looking back it reads like new improvements were arriving rapid-fire. Plus the patent on RSA expired in 2000 and it was set to...
Forum: Programming 2020-01-10, 02:52
Replies: 38
Views: 3,064
Posted By jasonp
Is Abramowitz and Stegun available anywhere...

Is Abramowitz and Stegun available anywhere online? This sounds like something they would have many possibilities available for.

In addition to the rational approximations, if you want simple code...
Forum: Msieve 2020-01-10, 02:32
Replies: 4
Views: 812
Posted By jasonp
Compute capability 2.0 is not supported in...

Compute capability 2.0 is not supported in latter-day versions of CUDA (9.1 and up I think). You also need to change the CUB compile line near the bottom of the Msieve makefile to remove the '200'...
Forum: Msieve 2019-12-09, 04:23
Replies: 4
Views: 590
Posted By jasonp
The code doesn't know the difference between...

The code doesn't know the difference between running in a directory with checkpoint files you don't want versus continuing a job with checkpoint files you do want. It would be a nice feature to...
Forum: Msieve 2019-12-09, 04:20
Replies: 72
Views: 10,332
Posted By jasonp
What card and CUDA version are you using? Do you...

What card and CUDA version are you using? Do you run the binary from command line, from the directory above the one with the PTX files?
Forum: Msieve 2019-11-07, 14:00
Replies: 14
Views: 1,292
Posted By jasonp
Yes, 1025 is the merge from the msieve-lacuda...

Yes, 1025 is the merge from the msieve-lacuda branch. The problem is likely not in the linear algebra per se but in the postprocessing that recovers the dependencies. Someday I will have to track it...
Forum: Msieve 2019-11-04, 15:21
Replies: 14
Views: 1,292
Posted By jasonp
If you are working with MPI over multiple...

If you are working with MPI over multiple machines, can you see if the postprocessing runs to completion using the latest Msieve svn? Greg has reported that it does not (the square root doesn't like...
Forum: Factoring 2019-11-01, 14:46
Replies: 10
Views: 886
Posted By jasonp
Tom is correct, Patrick talked about GPU block...

Tom is correct, Patrick talked about GPU block Wiedemann in late 2008 at the CADO conference. We never discussed his lattice sieve per se, I only discovered the code later.
Forum: Factoring 2019-10-22, 12:02
Replies: 11
Views: 838
Posted By jasonp
To be clear, I've read the paper and didn't see...

To be clear, I've read the paper and didn't see anything actionable. The problem with GPU SpMV is not that we don't know how to do one but that NFS matrices are so sparse that the GPU spends most of...
Forum: Factoring 2019-10-22, 00:05
Replies: 11
Views: 838
Posted By jasonp
Sparse matrix multiplies on GPU have become a lot...

Sparse matrix multiplies on GPU have become a lot more sophisticated since 2011 (incidentally $75 for an e-book is about the most expensive I have ever seen). See older versions of the moderngpu...
Forum: Factoring 2019-10-20, 20:39
Replies: 11
Views: 838
Posted By jasonp
There is a branch of Msieve that uses CUDA to...

There is a branch of Msieve that uses CUDA to implement block Lanczos on Nvidia GPUs. We were making steady progress before I had to put it down in 2013; the smaller size of GPU memory can be worked...
Forum: Msieve 2019-08-27, 15:18
Replies: 4
Views: 803
Posted By jasonp
The configuration for large polyselect jobs does...

The configuration for large polyselect jobs does many fairly outlandish things; I guess it's worth revisiting how configuration works so that we can get sensible defaults even without user-specific...
Forum: Soap Box 2019-08-02, 12:03
Replies: 16
Views: 481
Posted By jasonp
My understanding is that your insurance limit is...

My understanding is that your insurance limit is per account type, so 250k per person aggregated across all joint accounts at that institution for that person. Plus 250k for an account solely in that...
Forum: Forum Feedback 2019-07-30, 11:41
Replies: 12
Views: 1,338
Posted By jasonp
I lobbied for 'farang' in the text at the time....

I lobbied for 'farang' in the text at the time. Farang-utan?
Forum: Msieve 2019-07-30, 11:36
Replies: 15
Views: 1,983
Posted By jasonp
Congrats on a huge job complete! The...

Congrats on a huge job complete!

The postprocessing has needed a lot of hardening because we have seen all kinds of weird stuff in relation output lines. The one you posted also has a very large...
Forum: YAFU 2019-07-26, 19:52
Replies: 35
Views: 5,329
Posted By jasonp
I've been remiss in adding my congrats, getting...

I've been remiss in adding my congrats, getting end-to-end QS with three large primes is a major major undertaking.
Showing results 1 to 25 of 1000

 
All times are UTC. The time now is 03:48.

Sat Jul 11 03:48:08 UTC 2020 up 108 days, 1:21, 0 users, load averages: 1.64, 1.73, 1.56

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2020, 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.