mersenneforum.org  

Go Back   mersenneforum.org > Extra Stuff > Miscellaneous Math

Reply
 
Thread Tools
Old 2011-07-02, 15:43   #1
JohnFullspeed
 
May 2011
France

16110 Posts
Default Power 2

beforre to search can you please tell me if it's possible to find?

I need n with n=2 ^k and n product of primes (with 2 and without 5)
(Product of odd value by 2)
Thanks
JohnFullspeed is offline   Reply With Quote
Old 2011-07-02, 16:43   #2
Christenson
 
Christenson's Avatar
 
Dec 2010
Monticello

24×107 Posts
Default

John, you aren't making sense in English....any k will do for the first sentence, and the second one doesn't make sense. It's time we practised our French.
Christenson is offline   Reply With Quote
Old 2011-07-02, 17:54   #3
science_man_88
 
science_man_88's Avatar
 
"Forget I exist"
Jul 2009
Dumbassville

8,369 Posts
Default

Quote:
Originally Posted by Christenson View Post
John, you aren't making sense in English....any k will do for the first sentence, and the second one doesn't make sense. It's time we practised our French.
I don't trust his french either:

Quote:
bonjour

je ne suis pas sur que mon anglais soit assez bon pour dire a BSquared que le filtre qi'il utilise n'a pas duu tout les performances données (par l(auteur, pas par lui)
Pareillement je ne pense pas une seule seconde que l'erreur de l'auteur soit volontaire.
Le fioltre fonctionne, ses optimisationsaussi, mais la mesure n'est pas juste

En fait la methode a deplacée une partie du calcul hors chrono
On initialise LES tables des multiples avant alors que le crible d'Earsto le fait pendant
Il faut donc en toute justicce integré ce temps
à chaque mesure


Je n'ai pas peur de BSquared mais je ne voudrais surtout pas le vexerr et qu'il imagine que je ne repecte pas son travaiil. Seulemment cela m'interessait de savoir où je perdais du temps et j'ai trouve cette anmalie


Ne possédant pas un processeur CISC mais RISC je ne peux comparer le code avec le mien
Jérôme
google translate:

Quote:
hello

I'm not sure my English is good enough to say that the filter has BSquared qi'il duu not use any performance data (by (author, not by him)
Similarly I do not think for one second that the error of the author is voluntary.
The fioltre works, its optimisationsaussi, but the measure is not just

In fact the method has moved part of the computation off timer
It initializes the tables while before many of the screen for the fact Earsto
Therefore in all that time justicce integrated
each measure


I'm not afraid of BSquared but I would hate to imagine the vexerr and that I do not repect his travaiil. Seulemment I was interested to know where I lost time and I find this anmalie


Do not have a RISC processor CISC but I can not compare with my own code
Jerome

Last fiddled with by science_man_88 on 2011-07-02 at 17:55
science_man_88 is offline   Reply With Quote
Old 2011-07-02, 19:05   #4
Christenson
 
Christenson's Avatar
 
Dec 2010
Monticello

24×107 Posts
Default

SM88:
The french is significantly better than the English. Don't fully trust Google translations, especially in the presence of John's frequent typos -- here's what I get for the same, by hand:

I'm not sure my English is good enough to say to BSquared that the filter he uses doesn't at all have the stated[or claimed] performance (by the Author, not by him).
Similarly, I don't think for one instant that the author's error is intentional.
The filter works, as does its optimisations, but the metric is unfair.

In fact the method has moved part of the calculation off the clock,
It initializes THE tables of multiples beforehand while the Sieve is waiting.
In all fairness, this time must be integrated in every metric.

I'm not afraid of BSquared but above all I don't wish to vex him and have him think I don't respect his work. This interested me only in knowing where I was losing time and I found this anomaly.

Not possessing a CISC processor, only a RISc processor, I can't compare the code against my own.

*********
My Response to this:
I expect that the timing errors introduced by omitting initialising tables of multiples are quite small, probably smaller than those introduced by the presence of various operating systems such as Windows and Linux.

If you want to show respect, try programming the suggested optimisations.

Finally, most of the code is in high-level language, so it is quite portable. The parts that aren't, are either absolutely time-critical, or are doing things that can't be done in high-level language, such as letting the CPU know about how to predict branches. The guys here do these things because the programs they run, and have us run, take months to complete, so the small optimisations grow tremendously in importance.
Christenson is offline   Reply With Quote
Old 2011-07-02, 19:16   #5
JohnFullspeed
 
May 2011
France

7×23 Posts
Default power 2

en Français

j'ai besoin d'un nombre compose mais qui soit aussi une puissance de deux
Mais je pense que toutes les puissances de deux ne conaisse que 2 comme facteur
Jerome

( jai un souci avec mon clavier : la touche 'espace ne marche pas bien

The fioltre works, its optimisations aussi, but the measure is not just

Mais s'est de plus en plus dur.
JohnFullspeed is offline   Reply With Quote
Old 2011-07-02, 21:59   #6
fivemack
(loop (#_fork))
 
fivemack's Avatar
 
Feb 2006
Cambridge, England

2×29×109 Posts
Default

Bien sur, les puissances de deux ne comprime que 2 comme facteur.

Pour vos autres questions, il ne vaut vraiment le peine de contempler à quelle vitesse vous pouvez obtenir quelque chose en n'utilisant que quelque sous-ensembles des optimisations disponibles.

La methode TF ne trouve que des toutes petits facteurs (même a 10^8 c'est plus vite avec ECM), mais cela les trouve très vite; donc, on l'utilise toujours pour quelques secondes, et puis on continue avec des algorithmes de plus efficaces pour de plus grands facteurs.
fivemack is offline   Reply With Quote
Old 2011-07-02, 23:38   #7
Christenson
 
Christenson's Avatar
 
Dec 2010
Monticello

171210 Posts
Default

Quote:
Originally Posted by JohnFullspeed View Post
en Français

j'ai besoin d'un nombre compose mais qui soit aussi une puissance de deux
Mais je pense que toutes les puissances de deux ne conaisse que 2 comme facteur
Jerome

( jai un souci avec mon clavier : la touche 'espace ne marche pas bien

The fioltre works, its optimisations aussi, but the measure is not just

Mais s'est de plus en plus dur.
In French,

I'm looking for a composite but which is also a power of 2. But I think that all the powers of two only have 2 as a factor.

(I have a small problem with my keyboard: The space key isn't working

The filter works, its optimisations also, but the measurement isn't correct.

But it is becoming harder.

(I have
Christenson is offline   Reply With Quote
Old 2011-07-03, 07:14   #8
JohnFullspeed
 
May 2011
France

7×23 Posts
Default

Quote:
Originally Posted by fivemack View Post
Bien sur, les puissances de deux ne comprime que 2 comme facteur.

Pour vos autres questions, il ne vaut vraiment le peine de contempler à quelle vitesse vous pouvez obtenir quelque chose en n'utilisant que quelque sous-ensembles des optimisations disponibles.

La methode TF ne trouve que des toutes petits facteurs (même a 10^8 c'est plus vite avec ECM), mais cela les trouve très vite; donc, on l'utilise toujours pour quelques secondes, et puis on continue avec des algorithmes de plus efficaces pour de plus grands facteurs.

Je cherche a savoir cmment fonctionne TF pas pour comparer mais
poour trouver d'autres approches que lee miennes pour les adapter si possibles aux brnds nombres
Je ne sais meme pas a quoi sert TF se qui m'intresse c'est qu'il factorise

JJohn
JohnFullspeed is offline   Reply With Quote
Old 2011-07-03, 12:19   #9
Christenson
 
Christenson's Avatar
 
Dec 2010
Monticello

110101100002 Posts
Default

Quote:
Originally Posted by fivemack View Post
Bien sur, les puissances de deux ne comprime que 2 comme facteur.

Pour vos autres questions, il ne vaut vraiment le peine de contempler à quelle vitesse vous pouvez obtenir quelque chose en n'utilisant que quelque sous-ensembles des optimisations disponibles.

La methode TF ne trouve que des toutes petits facteurs (même a 10^8 c'est plus vite avec ECM), mais cela les trouve très vite; donc, on l'utilise toujours pour quelques secondes, et puis on continue avec des algorithmes de plus efficaces pour de plus grands facteurs.
Of course, the powers of 2 will only have 2 as a factor.

For your other questions, it is really worth the trouble of looking at what speed you can get something only using some subsets of the available optimisations.

The method of TF only finds very small factors (even beyond 10^8 its quicker with ECM), but it finds them very quickly, thus, it is always used for some seconds, and then it is followed with more effective algorithms for the larger factors.
Christenson is offline   Reply With Quote
Old 2011-07-03, 12:37   #10
Christenson
 
Christenson's Avatar
 
Dec 2010
Monticello

110101100002 Posts
Default

Quote:
Originally Posted by JohnFullspeed View Post
Je cherche a savoir cmment fonctionne TF pas pour comparer mais
poour trouver d'autres approches que lee miennes pour les adapter si possibles aux brnds nombres
Je ne sais meme pas a quoi sert TF se qui m'intresse c'est qu'il factorise

JJohn
I want to know how TF works not for comparison but
to find other approaches than mine to adapt them if possible to large numbers.

I don't even know what TF is, that which interests me is what it factors.

Reply:
TF = Trial Factoring = check sequentially all numbers that *could* factor a certain number do so. For example, if I want to factor 10,001, the method of TF starts with 2, discovers it doesn't factor 10,001, then 3(again, not a factor), then 5, then 7, then 11,....up to 100 or until it finds a number that divides 10,001.

There are optimisations on Mersenne numbers that can TF to 2^50 in just a few seconds, and require fewer factors checked below a given limit as the exponent grows larger.

http://www.mersenne.org/various/math.php explains some of this.
Christenson is offline   Reply With Quote
Old 2011-07-03, 15:16   #11
JohnFullspeed
 
May 2011
France

101000012 Posts
Default Merci

Thanks
John
JohnFullspeed is offline   Reply With Quote
Reply

Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
More Power!!!! petrw1 Teams 10 2019-10-15 17:36
TDP as power used? CRGreathouse Hardware 9 2016-02-06 18:46
How much power am I really using? petrw1 Lounge 19 2013-12-13 13:00
Power??? JohnFullspeed Programming 5 2011-08-30 16:28
IBM Power 6 Unregistered Information & Answers 7 2008-08-30 14:36

All times are UTC. The time now is 03:34.

Mon Oct 26 03:34:48 UTC 2020 up 46 days, 45 mins, 0 users, load averages: 2.05, 2.06, 1.93

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.