mersenneforum.org Search Results
 Register FAQ Search Today's Posts Mark Forums Read

 Showing results 1 to 25 of 81 Search took 0.02 seconds. Search: Posts Made By: Cyclamen Persicum
 Forum: Software 2006-12-31, 15:03 Replies: 1 Views: 1,030 Posted By Cyclamen Persicum GWNUM as DLL? Is there GWNUM as precompiled DLL, just to easy call fast multiplication and squaring? I have tested new version of PRP.EXE (24.14) and found that it incredibly, fantastically fast, 20 times...
 Forum: Math 2004-05-15, 07:51 Replies: 23 Views: 1,960 Posted By Cyclamen Persicum A 3000 digit long prime number is not a problem... A 3000 digit long prime number is not a problem at all =))). My comp has worked half an hour and gave me the following 10240-bit long primes: 0835h*2^10228 + 1 355Fh*2^10226 + 1 48F9h*2^10225...
 Forum: Math 2004-05-14, 14:47 Replies: 23 Views: 1,960 Posted By Cyclamen Persicum I use the GMP-Routine to perform the modular... I use the GMP-Routine to perform the modular power calculation. This is pretty fast. It takes about 5 to 10 seconds on a Pentium III machine (666 Mhz) for a number which is about 3000 digits long. ...
 Forum: Math 2004-05-14, 04:45 Replies: 23 Views: 1,960 Posted By Cyclamen Persicum You have not said the main thing! How long... You have not said the main thing! How long prime are you looking for? How fast is your Modular Power? Is it as fast as mine (if you downloaded my prog), or it is as fast as PRP.EXE, or you don't...
 Forum: Math 2004-05-13, 13:28 Replies: 23 Views: 1,958 Posted By Cyclamen Persicum (10^p)-1 is never prime You are right... ... (10^p)-1 is never prime You are right... :spam: But we are talking about primes (10^p-1)/9, i.e. about numbers which have only 111111... in their decimal representation. It is 2 units(...
 Forum: Math 2004-05-13, 10:32 Replies: 23 Views: 1,958 Posted By Cyclamen Persicum Finally, your use of the term "Generalized... Finally, your use of the term "Generalized Mersenne Prime" is a mis-nomer. Mathematicians have used the term for years, but it means something entirely different. Note that a Mersenne prime is a...
 Forum: Math 2004-05-13, 10:24 Replies: 23 Views: 1,960 Posted By Cyclamen Persicum How do you perform your trial division? You can... How do you perform your trial division? You can use modular arithmetic or sieve technique. Let your number has a form: k*p+1, you can calculate only once z=p mod s, then k*p+1 = k*z+1 (mod s)....
 Forum: Hardware 2004-05-09, 06:01 Replies: 4 Views: 1,167 Posted By Cyclamen Persicum ASM Optimization I have downloaded Intel Manual - Pentium Instruction Set vol.2A-2B, files 25366613.pdf and 25366713.pdf How can I find info about number of ticks (or tacts) for every ASM instruction? One my...
 Forum: Miscellaneous Math 2004-05-08, 05:10 Replies: 29 Views: 2,260 Posted By Cyclamen Persicum I see what you want to say: Instead... I see what you want to say: Instead (x*x*x*x*x*x*...*x) mod y it is possible do not calculate the whole huge product of x, and one should take modulo after every mult, ((((x*x) mod y) *x) mod y)...
 Forum: Math 2004-05-05, 15:32 Replies: 23 Views: 1,960 Posted By Cyclamen Persicum No I wrote this program myself. What is a... No I wrote this program myself. What is a PRP-Program by the way? and what does the program do you mentioned? That is great!!! I also have written a probable-prime prog:...
 Forum: Math 2004-05-04, 14:41 Replies: 23 Views: 1,960 Posted By Cyclamen Persicum Have you written your own PRP prog or you just a... Have you written your own PRP prog or you just a user of NewPGEN-PRP.EXE ?
 Forum: Math 2004-05-03, 16:35 Replies: 23 Views: 1,960 Posted By Cyclamen Persicum It doesn't need to be totally accurate. You... It doesn't need to be totally accurate. You need not primality test. You need a "compositeness" test, i.e. kinda test which never miss if it says "composite". It can say "dont' knoff" but if it...
 Forum: Math 2004-05-01, 14:39 Replies: 23 Views: 1,960 Posted By Cyclamen Persicum I guess it is much faster, isn't it? No. When I... I guess it is much faster, isn't it? No. When I said "Rabin test" I meant it as a synonym to Fermat test, although Rabin is STRONG pseudoprime test but Fermat itself is weak and pseudoprime...
 Forum: Math 2004-05-01, 05:23 Replies: 23 Views: 1,960 Posted By Cyclamen Persicum I have already asked about it: "Is there kind... I have already asked about it: "Is there kind of "middle" test, much slower than trial division on 3,5,7 but still much faster then Rabin??? I want to embed it before Rabin-test." But I didn't...
 Forum: Math 2004-04-24, 18:00 Replies: 7 Views: 5,247 Posted By Cyclamen Persicum By the way, I have found an advantage of Prothes... By the way, I have found an advantage of Prothes in sence of speed. When performing destributed calculations, all 10,000,000 digit prothes will require the same time to check. And in the case of...
 Forum: Math 2004-04-24, 11:56 Replies: 28 Views: 2,624 Posted By Cyclamen Persicum Golden Intersection = 2*cos(pi/5) Golden Intersection = 2*cos(pi/5)
 Forum: Hardware 2004-04-24, 11:00 Replies: 11 Views: 1,361 Posted By Cyclamen Persicum I'm looking at improving the FFT routines... ... I'm looking at improving the FFT routines... How will I know about new version of PRP.EXE
 Forum: Math 2004-04-24, 10:29 Replies: 7 Views: 5,247 Posted By Cyclamen Persicum Here you are, Bob Silverman!!! Nice to hear... Here you are, Bob Silverman!!! Nice to hear you, an angry man, an implacable enemy of bullshit, a preceptor of erringers. Are you one of a couple of tens so called GIMPS experts? That is an...
 Forum: Hardware 2004-04-23, 16:04 Replies: 11 Views: 1,361 Posted By Cyclamen Persicum Thank you very much for your answer. You are the... Thank you very much for your answer. You are the greatest programmer of all times and among all people indeed. I appreciate your attention. Currently I have written two progs which deal with long...
 Forum: Math 2004-04-23, 05:04 Replies: 7 Views: 5,247 Posted By Cyclamen Persicum Why so complex? You need a 10 million digit... Why so complex? You need a 10 million digit prime? It is easy. Let's N=k*2^33,000,000+1 Then choose small k and a for which a^[(N-1)/2]=-1 mod N Thats all falks. And it is easier than...
 Forum: Hardware 2004-04-22, 13:02 Replies: 11 Views: 1,361 Posted By Cyclamen Persicum That is an impressive accomplishment for high... That is an impressive accomplishment for high level language programming. The secret is that I call fftw3.dll from fftw.org, and after performing an inverse FFT all I need is to take care about...
 Forum: Hardware 2004-04-21, 15:51 Replies: 11 Views: 1,361 Posted By Cyclamen Persicum Hey, was perfectly you-Prime95 who wrote PRP.EXE... Hey, was perfectly you-Prime95 who wrote PRP.EXE ? It works two times faster than my prog with 256Kbit number. 900s vs. 1900s =((( This is due to such kinda cach tricks and ruses... On the other...
 Forum: Math 2004-04-20, 04:54 Replies: 1 Views: 687 Posted By Cyclamen Persicum Sorry, I have noticed a pun... JacobiSymbol =... Sorry, I have noticed a pun... JacobiSymbol = -1 does not mean "quadratic non-residue". But I'm right, there is a super-extended Proth theorem: for integer _a_ with Jacobi(a,N)=-1 (or for odd _a_...
 Forum: Math 2004-04-19, 12:35 Replies: 1 Views: 687 Posted By Cyclamen Persicum Extended Proth Theorem I am writing a Proth prime generator (yet another... ;-) The Proth theorem says: "Let n > 1, k < 2^n and N = k*2^n + 1 be a quadratic non-residue (mod a) for some odd prime a. Then the necessary...
 Forum: Miscellaneous Math 2004-04-16, 14:21 Replies: 17 Views: 3,427 Posted By Cyclamen Persicum the L-L primality test that works for Mersenne... the L-L primality test that works for Mersenne numbers is much faster than tests that work for GF numbers. I dont' think so. I believe that all classical N +/- 1 tests have the same laboriousness....
 Showing results 1 to 25 of 81

All times are UTC. The time now is 19:10.

Sat Nov 28 19:10:41 UTC 2020 up 79 days, 16:21, 3 users, load averages: 1.46, 1.53, 1.57