mersenneforum.org

mersenneforum.org (https://www.mersenneforum.org/index.php)
-   Software (https://www.mersenneforum.org/forumdisplay.php?f=10)
-   -   New Mersenne Software For Test Mersenne Prime Numbers On Android (https://www.mersenneforum.org/showthread.php?t=23963)

thorken 2019-01-03 15:15

New Mersenne Software For Test Mersenne Prime Numbers On Android
 
4 Attachment(s)
Hello first of all tell you that looking for things about prime numbers I have found this website I am Spanish and I found on a mobile that I found this application in the folder / system / xbin of the phone so if you want to try it you will have to be root users and copy it to this folder once copied you give permissions as other Apps that are in that folder and ready, then you open a terminal as root and you write mersenne press enter and the application is already running. I've been doing some test and it calculates faster than the Google Play app Luke Lehmer. Basically the app gives you a range of exponents and calculates the found mersenne cousins ​​and saves them in a file also if you only want to analyze 1 with the same exponent in the interval that the program asks for, because it analyzes it and that's it, I I have this program in an S4 Mini that I found and upload it here so we can use it all. Sorry for my English, if anyone has any questions you can ask me, I forgot in the xbin mask of this mobile I also found other programs of prime numbers and pi digits and very interesting so if someone wants to upload them tell me , at all times I upload only the mersenne binary. I upload some screenshot of the program for the distrustful ones. In the program is the email of the author everything is that in the end I claim the phone haha. Well I upload it. the download link for the app is here [url]https://drive.google.com/file/d/1-Tz7f6gZCSLhLabJGtJ_-9aBo_jWGf7-/view?usp=sharing[/url]

Batalov 2019-01-03 15:58

[COLOR=DarkRed]Don't upload blind binaries here. Period.[/COLOR]
You can upload screenshots, that's fine.

Also, this idea is >10 years too late. People have run (a much more efficient code) before. See here - [URL]https://mersenneforum.org/showthread.php?t=10691[/URL]

thorken 2019-01-03 16:02

entonces como subo la aplicacion para que la use la gente
 
then the app not upload ?

Batalov 2019-01-03 16:03

Publish the code. Don't expect people to root their phones for the binary app that does unknown damage.

thorken 2019-01-03 16:05

this app its mor fast than you saw me
 
this app is more faster that you saw me

thorken 2019-01-03 16:07

could i upload the app to google drive and put the link to download ?

jnml 2019-01-03 16:38

[QUOTE=thorken;504770]could i upload the app to google drive and put the link to download ?[/QUOTE]

Plase do not do that even if allowed here. No one sane should download the binary and run it
as root on his/her phone. Actually, no binary at all, not only this one, from non trusted
source.

thorken 2019-01-03 16:43

the app its good
 
i upload the screenshots and i run the app in my phone its good and not problem,on the other hand that executes the application the one that wants does not have major problem

thorken 2019-01-03 16:54

M86243 In 3 minutes 7 seconds on s4 mini gt i9195
 
2 Attachment(s)
M86243 In 3 minutes 7 seconds on s4 mini gt i9195
M23209 in 6 seconds on s4 mini gt i9195

thorken 2019-01-03 17:20

code more eficcient where ?
 
[QUOTE=Batalov;504764][COLOR=DarkRed]Don't upload blind binaries here. Period.[/COLOR]
You can upload screenshots, that's fine.

Also, this idea is >10 years too late. People have run (a much more efficient code) before. See here - [URL]https://mersenneforum.org/showthread.php?t=10691[/URL][/QUOTE]

i dont found any app for android more efficent that this you not probe the app yet? when you probe the app talk.

science_man_88 2019-01-03 17:34

[QUOTE=thorken;504781]i dont found any app for android more efficent that this you not probe the app yet? when you probe the app talk.[/QUOTE]

The point is we don't have reason to trust code that isn't open source. In VB.net I theoretically could make a virus of 6 lines and hide it in a real looking app. If I want an app I'll use Paridroid ( PARI/GP for android).

kriesel 2019-01-03 17:45

A prudent person will hesitate to run random software, even forwarded to them by someone they've known all their lives. One of my sisters used to forward me things in email like elfbowl.exe, that she got from a friend, who got it from a friend, etc. She thought I was being unreasonable and paranoid and resistant to sharing the harmless fun, when I asked her to stop sending them, as I had no intention of running them. Then one day she had to pay the local computer store to reinstall her operating system and everything else, except her personal data that was gone forever.
When one of the requirements is to run the suspect software from an unknown source as root, extra skepticism is justified.

thorken 2019-01-03 18:51

[QUOTE=science_man_88;504784]The point is we don't have reason to trust code that isn't open source. In VB.net I theoretically could make a virus of 6 lines and hide it in a real looking app. If I want an app I'll use Paridroid ( PARI/GP for android).[/QUOTE]


then as you trust the code that is out there or the apps that go up to google play, this is not any virus that something happens to me that denounces me is reliable code.

thorken 2019-01-03 18:55

M216091 in 23min 7 sec
 
1 Attachment(s)
M216091 in 23min 7 sec in s4 mini gt-i9195

GP2 2019-01-03 18:56

[QUOTE=thorken;504778]M86243 In 3 minutes 7 seconds on s4 mini gt i9195[/QUOTE]

[QUOTE=thorken;504794]M216091 in 23min 7 sec in s4 mini gt-i9195[/QUOTE]

By modern standards that is extremely slow. These small exponents should take seconds, not minutes.

thorken 2019-01-03 19:01

[QUOTE=kriesel;504786]A prudent person will hesitate to run random software, even forwarded to them by someone they've known all their lives. One of my sisters used to forward me things in email like elfbowl.exe, that she got from a friend, who got it from a friend, etc. She thought I was being unreasonable and paranoid and resistant to sharing the harmless fun, when I asked her to stop sending them, as I had no intention of running them. Then one day she had to pay the local computer store to reinstall her operating system and everything else, except her personal data that was gone forever.
When one of the requirements is to run the suspect software from an unknown source as root, extra skepticism is justified.[/QUOTE]

the app have been executed in my three moviles without problems its a program to calculate mersenne primes nothing more... and its make it work very good, if dude execute in emulator android in virtual machine.

thorken 2019-01-03 19:02

[QUOTE=GP2;504795]By modern standards that is extremely slow. It shouldn't take more than one second for such a small exponent.[/QUOTE]

this program run in android, in android i dont know a app that run more faster that this. if you know any app that runs faster say me please.

GP2 2019-01-03 19:06

[QUOTE=thorken;504799]this program run in android, in android i dont know a app that run more faster that this. if you know any app that runs faster say me please.[/QUOTE]

Use a PC instead. The mprime/Prime95 program is the fastest.

This is not the kind of program you should run on a phone, it will overheat and really drain your battery.

Batalov 2019-01-03 19:15

A smartphone is not really a smartphone anymore if it needs to be plugged into the power supply all the time. And if you don't plug it - it will drain your battery in minutes. And you can overheat and some sloppily connected parts inside your phone will unglue - and what are you going to do next, pray tell?

More importantly - how are you going to compensate gullible people who will download your app and to whom it will happen? At best you will just say 'sorry' over the internet. That's gonna make them all warm and fuzzy, for sure.

paulunderwood 2019-01-03 19:20

[QUOTE=thorken;504794]M216091 in 23min 7 sec in s4 mini gt-i9195[/QUOTE]

Try testing this one which is on the wave front of GIMPS testing: M82589933.

How long will it take? Does it fit in the RAM of your smart phone?

thorken 2019-01-03 19:22

[QUOTE=GP2;504800]Use a PC instead. The mprime/Prime95 program is the fastest.

This is not the kind of program you should run on a phone, it will overheat and really drain your battery.[/QUOTE]

in this forum talk in a post about people that want a app to calculate mernenne primes in android on mobile phones and i posted the app. a mobile phone in 2019 its more powerfull that a pc about 10 years and most people like to probe apps of this type in your mobiles. i run the app for more that ten hours and the temp of the cpu nots up of 62g centigrate. obviusly plug to the charger. and remember that your phone its more powerfull that computers of 10 years ago or minus.

thorken 2019-01-03 19:25

[QUOTE=paulunderwood;504803]Try testing this one which is on the wave front of GIMPS testing: M82589933.

How long will it take? Does it fit in the RAM of your smart phone?[/QUOTE]

run on dual core or in pentium 4 or in i3. and this app dont problems of memory for larger numbers. its good program. and for test large mersene numbers dont require a big ram. only big compute power.

Prime95 2019-01-03 19:26

[QUOTE=thorken;504799]this program run in android, in android i dont know a app that run more faster that this. if you know any app that runs faster say me please.[/QUOTE]

First off, welcome to the forum.

Congratulations on creating a working Lucas-Lehmer test program. I hope you have found the experience enjoyable and informative.

The skepticism here is no reflection on you personally. There are two reasons your announcement has not received the enthusiasm you hoped for. 1) The processor in a phone is not powerful enough to contribute to the search for new Mersenne primes. 2) People here are reluctant to download *any* software from the Internet.

Instead, I suggest posting your code. Members here can suggest improvements or marvel at your ingenuity.

thorken 2019-01-03 19:28

[QUOTE=paulunderwood;504803]Try testing this one which is on the wave front of GIMPS testing: M82589933.

How long will it take? Does it fit in the RAM of your smart phone?[/QUOTE]

in play store a app called lucas lehmer its for android very very slow. this app its about twenty times more faster.

science_man_88 2019-01-03 19:36

[QUOTE=thorken;504807]in play store a app called lucas lehmer its for android very very slow. this app its about twenty times more faster.[/QUOTE]

Okay, how does it perform in Big O notation ?

[url]https://rob-bell.net/2009/06/a-beginners-guide-to-big-o-notation/[/url]

kriesel 2019-01-03 19:44

[QUOTE=thorken;504798]the app have been executed in my three moviles without problems its a program to calculate mersenne primes nothing more... and its make it work very good, if dude execute in emulator android in virtual machine.[/QUOTE]

How many of the criteria does it pass? How can we know, independent of taking the word of a stranger or gambling the health of our personal phones? (I'm not inclined to buy a burner phone to test it out myself, because I know from the power budget of a phone its throughput will not be competitive.) [URL]https://www.mersenneforum.org/showpost.php?p=488298&postcount=4[/URL]

What are its exponent limits versus phone memory size or other parameters? [URL]https://www.mersenneforum.org/showpost.php?p=488291&postcount=2[/URL]

kriesel 2019-01-03 20:18

[QUOTE=thorken;504758]Hello ...[/QUOTE]
Since my grasp of Spanish is almost nonexistent, I keyed in an American-English-keyboard approximation of the first screen shot, ran it through Google Translate, and got:
[CODE]MvIiIaX - Mersenne Cousins ​​Numbers Calculation - MvIiIaX
The Mersenne Prime Numbers, Are Numbers Cousins ​​Such That ... 2 ^ n-1 is a Prime Number, n Being Cousin. Example: If We Give An Interval From 0 To 10000 When n Be 9973 We Will Obtain A Number Of Some 3003 Digits To Analyze And See If It Is A Prime Number Or No. The Mersenne Prime Numbers Are Represented As: M2, M3, M5, M7, M13, M17, M19 ... M82589933 That's the 51st Mersenne Prime Number Calculated By January 2019, Which Has More Than 24 Million Digits, Much Much Calculation. The purpose of this program is mathematical and Benchmark mode in a given interval, for these calculations it takes much more power than a mobile gives to January 2019, although you will know ... within a few years this program to be surprised ...

At the end of the calculation, the program will give us a rate of volume, this rate of speed of your mobile generating cousins ​​of Mersenne, will have a maximum of 100 points, the higher you are, the more powerful your mobile, for example a S4 Mini Year 2013, Gives An Index Of 0.176 Points, Calculating From 0 To 10000 Of Exponent,? And yours ?. You Can Serve In Way Of BenchMark In Your Mobile Equipment Or To Compete With Your Friends ... To See Who Has The Best Mobile. This Program Does Intensive Use Of The CPU, So The Quicker It Is, The Quicker The Calculations ... The File M8AX - Mersenne.TXT Will Be Created With The Results Of The Calculations In The Folder In Which You Execute The Program.

Enter the Exponent from which the calculations will start.
0

Enter The Exponent In Which The Calculations Will End.
10000[/CODE]


Anybody know what "MvIiIaX" is?

pinhodecarlos 2019-01-03 20:38

I thought your second language learnt at school was Spanish.

science_man_88 2019-01-03 20:51

[QUOTE=kriesel;504816]
Anybody know what "MvIiIaX" is?[/QUOTE]

I don't know but it seems close to melilax ( appears to be a honey based enema)

kriesel 2019-01-03 21:00

Run time scaling from Thorken's posts
 
1 Attachment(s)
[QUOTE=pinhodecarlos;504817]I thought your second language learnt at school was Spanish.[/QUOTE]German. Before PCs were a thing. Just a touch of Spanish for travel purposes (please, thank you, where is the bathroom, etc)

~p[SUP]2.45[/SUP], about a century for M(82859933) from a very long extrapolation made from only 3 data points and limited precision.

What I've seen for other codes range from 2.03 up. 2+ is the minimum, One for iteration count, one for operand size, plus a bit of change for fewer bits/word in the fft transform at larger sizes. Or n * n log n log log n. CUDALucas 2.094, prime95 2.094. [URL]https://www.mersenneforum.org/showpost.php?p=502778&postcount=2[/URL]
[URL]https://www.mersenneforum.org/showpost.php?p=488523&postcount=2[/URL]


(Folks in the know, don't take that as talking down to you. It's for those looking up at the learning curve still ahead.)

[QUOTE]Instead, I suggest posting your code.[/QUOTE]Meaning, source code.

firejuggler 2019-01-03 21:13

no need for 'burner' phone.. there are android emulator out there...( well for windows at least)
Bluestack, Memu,LDplayer, Andy and Nox.

Since I have a dumbphone, if I want to play some mobile game, I have to use an emulator.

I personnally prefer bluestack.

ewmayer 2019-01-03 22:13

[QUOTE=thorken;504804]a mobile phone in 2019 its more powerfull that a pc about 10 years[/QUOTE]

That may be so, but in that case your program is not making very good use of your hardware, because on my 2009-vintage Core2Duo Macbook classic I can do the LL test of 216091 an order of magnitude faster than the timing you posted.

I also notice that your runtime for 216091 is more than 7x that for 86243, which is somewhat above the O(n^2 log n) scaling one expects from a fast-transform-based large-integer-multiply algorithm. What algorithm are you using for the modular mutiply needed by the LL test?

ATH 2019-01-03 22:32

[QUOTE=thorken;504794]M216091 in 23min 7 sec in s4 mini gt-i9195[/QUOTE]

The smallest exponent needed for the project is around M47,000,000 for double checking.

If it takes 23 min for M216091 it will take months / years for those exponents if the phone can handle them at all.


Also consider that this project have been running for 23 years! since january 1996. It is very unlikely anyone just joining the project can come up with something ground breaking that no one else thought about in all those years.

A few very intelligent and clever people have made huge changes and contributions to this project over the years, but I do not think any of them had just joined when they did it.

To quote an old man you may or may not know: "Stay awhile and listen..."

axn 2019-01-04 02:41

How difficult will it be to compile MLucas for android ?

I would love to see the performance of an actual state-of-the-art code running on one of these.

LaurV 2019-01-04 06:53

@OP: can you give us some info about the multiplication algorithm used? Is it school grade, is it karatsuba? FFT?

As per firejuggler's suggestion, we tried to run this toy in an android emulator - our kingdom is Cortex M, but we have some android emus laying around, as we have colleagues who develop applications with them, for our customers. We wanted to see how the tests scale in time - if the application is bogus, like reading exponents from a list and saying prime/composite, and doing other things meantime, then the time-scaling would be also odd. We have some idea and could guess the multiplication algorithm used, if any tuning was done or it is just blind school-grade high-level stuff. And we consider this to be a good test, because it would not be easy for the guy to fake the process (which involves exponentiations and multiplications related not only to the size of the exponents, but also to the number of bits which are 1 involved). And our toys can also debug/disassemble code. But unluckily, we can not run it. Our emulator does not like it, and our colleagues who could help didn't come back from their NY holidays yet...

Our advice for now: avoid it.

P.S. Carlos, you seem to be the only one liking it? Can you detail why? Or was that a voting mistake? :razz:

pinhodecarlos 2019-01-04 07:11

Being Portuguese I’m being supportive with our Spanish brother but also I’m affected by flu since new year on my leave. Call me crazy: am I tired of fights with Spanish or is flu damaging my brain?!??? Or both?!

LaurV 2019-01-04 07:22

Ok, that makes sense! (being supportive). :tu:

GP2 2019-01-04 09:29

Here is a Mersenne prime test. I didn't write it. It uses the GMP library.

It prints a 64-bit hexadecimal residue: "0" for a Mersenne prime, non-zero otherwise.

I compiled on Linux and ran on a single Skylake core on AWS.

For 86243 it takes 11 seconds; for 216091 it takes 1 minute 38.5 seconds.

And yet this is much slower than mprime/Prime95.

[CODE]
/* Adapted from http://rosettacode.org/wiki/Lucas-Lehmer_test#GMP */

#include <stdio.h>
#include <stdlib.h>
#include <gmp.h>

void lucas_lehmer(unsigned long p, mpz_t *V)
{
mpz_t mp, t;
unsigned long k;

mpz_init_set_ui(t, p);

mpz_init(mp);
mpz_setbit(mp, p);
mpz_sub_ui(mp, mp, 1);

mpz_init_set_ui(*V, 4);
for (k = 3; k <= p; k++) {
mpz_mul(*V, *V, *V);
mpz_sub_ui(*V, *V, 2);
/* mpz_mod(*V, *V, mp) but more efficiently done given mod 2^p-1 */
if (mpz_sgn(*V) < 0) mpz_add(*V, *V, mp);
/* while (n > mp) { n = (n >> p) + (n & mp) } if (n==mp) n=0 */
/* but in this case we can have at most one loop plus a carry */
mpz_tdiv_r_2exp(t, *V, p);
mpz_tdiv_q_2exp(*V, *V, p);
mpz_add(*V, *V, t);
while (mpz_cmp(*V, mp) >= 0) mpz_sub(*V, *V, mp);
}

mpz_clear(t); mpz_clear(mp);

/* Residue is returned in V */
}

int main(int argc, char* argv[]) {
mpz_t m64, V;

/* We want to print a 64-bit residue */
mpz_init(m64);
mpz_setbit(m64, 64);
mpz_sub_ui(m64, m64, 1);

unsigned long p;
if (argc >= 2) {
p = strtoul(argv[1], 0, 10);
}
else {
fprintf(stderr, "Provide one argument, a Mersenne exponent, for instance: 11213\n");
return 1;
}

lucas_lehmer(p, &V);

mpz_and(V, V, m64);
mpz_out_str(stdout, 16, V);
fprintf(stdout, "\n");

mpz_clear(m64); mpz_clear(V);

return 0;
}
[/CODE]

LaurV 2019-01-04 09:48

[QUOTE=LaurV;504878]Ok, that makes sense! (being supportive). :tu:[/QUOTE]
Please don't do this to my posts! Now there are two of you, and I don't know who to blame! :rant:
(now I know why Xyzzy gave the knife to Batalov too! :davar55:)

Batalov 2019-01-04 15:53

[QUOTE=thorken;504778]M86243 In 3 minutes 7 seconds on s4 mini gt i9195
M216091 in 23min 7 sec in s4 mini gt-i9195
[/QUOTE]
[QUOTE=GP2;504895]Here is a Mersenne prime test. I didn't write it. It uses the GMP library.

It prints a 64-bit hexadecimal residue: "0" for a Mersenne prime, non-zero otherwise.

I compiled on Linux and ran on a single Skylake core on AWS.

For 86243 it takes [COLOR=green][B]11 seconds[/B][/COLOR]; for 216091 it takes [COLOR=Green][B]1 minute 38.5 seconds.[/B][/COLOR]
[/QUOTE]

Great job, Gord! :tu:

axn 2019-01-04 16:28

3 min 7 sec / 11 sec = 17x
23min 7 sec / 1 min 38.5 sec = 14x

OP's code has better scaling than GMP-based one !

Mysticial 2019-01-04 17:25

[QUOTE=axn;504938]3 min 7 sec / 11 sec = 17x
23min 7 sec / 1 min 38.5 sec = 14x

OP's code has better scaling than GMP-based one ![/QUOTE]

Granted, GMP isn’t a very high bar to beat. Their refusal to use floating point means they don’t use any fast algorithms until much larger sizes.

science_man_88 2019-01-04 17:34

[QUOTE=axn;504938]3 min 7 sec / 11 sec = 17x
23min 7 sec / 1 min 38.5 sec = 14x

OP's code has better scaling than GMP-based one ![/QUOTE]

by a factor of roughly 1/2< ratio of n's > if my math is correct, but it in theory takes until exponents of 2.9 to 3 million to surpass it.

kriesel 2019-01-04 19:21

Fixed overhead can give lower order empirical timing scaling from low p measurements and make an implementation look like it's using a better algorithm than it is. In other words, the order bends upward to higher power of p at higher exponent, as setup overhead gets diluted by longer runtimes for higher exponent. This can be seen in runtime scaling measurements and plots I made for cllucas, CUDALucas and gpuowl. The more setup overhead, the lower the apparent order at low p. OpenCL's compile on the fly guaranteed around 2 seconds of setup overhead at each fft length. CUDALucas scaling: [URL]https://www.mersenneforum.org/showpost.php?p=488523&postcount=2;[/URL] gpuowl V5 scaling: [URL]https://www.mersenneforum.org/showpost.php?p=502776&postcount=10;[/URL] prime95 scaling: [URL]https://www.mersenneforum.org/showpost.php?p=502778&postcount=2[/URL] The per-iteration time scaling of gpulucas, cllucas, CUDALucas, and gpuowl v1.9 are compared at [URL]https://www.mersenneforum.org/showpost.php?p=488476&postcount=8[/URL]

I'd expect grammar-school multiplication based primality testing to scale as ~p[SUP]3[/SUP]. With sufficient overhead, it may scale as p[SUP]2.43[/SUP] at the low end, as in my ancient inefficient primitive algorithm code from the late 80s and early 90s does at low p. (There, inefficient TF is part of the run time, as is printout of every iteration's res64 and very frequent state saves for resumption as insurance against the vagaries of MS-DOS and no UPS.) Thorken's timings indicate scaling of p[SUP]2.42[/SUP], very close to that. It could be grammar school, or something better.

ewmayer 2019-01-04 20:38

[QUOTE=axn;504852]How difficult will it be to compile MLucas for android ?

I would love to see the performance of an actual state-of-the-art code running on one of these.[/QUOTE]

What kind of hardware CPU hardware are recent-model Android phones based on - Arm Cortex?

Again, no one in their right mind uses their actual phone to test M-numbers, I'm thinking more along the lines of a "buy a bunch of cheap used phone mainboards, hook to some kind of barebones power/comms harness, slap a little heatsink on each CPU, crunch away" kind of setup. No idea if that could potentially provide more bang-for-the-buck than, say, an Odroid micro-PC cluster, but surely there are many millions of such phones which get retired due to obsolescence, screen-damage, etc every year, but for which the mainboard is still functional. These things could potentially offer a cost-competitive LL-crunching solution, but one needs a lot of them (perhaps 20 or so Cortex A53 quads to equal a high-end Intel quad CPU) to match the crunching power, and so they need to be cheap. In the wilder of my fever dreams I picture 50-100 retired-phone mainboards bought on the cheap and assembled into a single chassis with similar footprint to a rackmount server blade a few inches high. :)

kladner 2019-01-04 21:16

1 Attachment(s)
[QUOTE=ewmayer;504953]What kind of hardware CPU hardware are recent-model Android phones based on - Arm Cortex?[/QUOTE]
[URL]https://www.phonearena.com/news/Samsung-Galaxy-S8-benchmarks_id93192[/URL]
[QUOTE]Like has happened so many times before, Samsung's built two main versions of the Galaxy S8 (and correspondingly, GS8+), one running one of the company's own Exynos chips, and one featuring a top-of-the-line Qualcomm Snapdragon processor. And what makes that pairing especially exciting this time around is that this is the very first time we've been able to interact with either chip on a commercial handset.

Samsung Galaxy S8 processor showdown: Exynos 8895 vs Snapdragon 835
In most international markets, the Galaxy S8 is built around the new Exynos 8895, featuring new custom CPU cores, new Mali GPU, and with the whole thing fabricated via a cutting-edge 10nm process intended to maximize performance and power efficiency.[/QUOTE]Based on the model number reported by my phone, a Galaxy S8, the zip file should be the correct specs. Model # is SM-G950U1. The CPU is a Samsung 8 core, with fast and slow quad units.

ewmayer 2019-01-04 22:01

[QUOTE=kladner;504957][URL]https://www.phonearena.com/news/Samsung-Galaxy-S8-benchmarks_id93192[/URL][/QUOTE]

Thanks - that glosses over the processor details, but that info is easily found - Here is [url=https://www.anandtech.com/show/11540/samsung-galaxy-s8-exynos-versus-snapdragon]AnandTech[/url]:
[quote]Our initial look at Snapdragon 835 revealed that its Kryo 280 performance cores are loosely based on ARM’s Cortex-A73 while the efficiency cores are loosely based on the Cortex-A53. Samsung's Exynos 8895 also has an octa-core big.LITTLE CPU configuration, but uses four of its own custom M2 cores paired with four A53 cores. Samsung introduced its first custom CPU core, the M1, last year. Compared to ARM’s A72, integer IPC was similar but the M1 trailed the A72 in efficiency. The M2 does not appear to be a radical redesign, but rather a tweaked M1 that offers the usual promises of improved performance and efficiency. Are the changes enough to top Qualcomm’s flagship SoC?[/quote]
That doesn't say precisely what "loosely based on" means - I'm hoping that at least the basic instruction set is identical.

By way of experimentation, we really need just one phone containing such an Arm-quad-based CPU and a suitable developer-tools interface of the kind app-writers use. In our case, said dev-tools could actually be just barebones gcc+libs.

thorken 2019-01-05 00:48

Run in a mobile not in i7
 
[QUOTE=Batalov;504931]Great job, Gord! :tu:[/QUOTE]

The app its for mobile phone I make it for mobile phones and for mobile phones I dont show a app more faster than this only for mobile.

science_man_88 2019-01-05 01:06

[QUOTE=thorken;504973]The app its for mobile phone I make it for mobile phones and for mobile phones I dont show a app more faster than this only for mobile.[/QUOTE]

which algorithm of LL testing do you use ?

thorken 2019-01-05 01:08

[QUOTE=LaurV;504875]@OP: can you give us some info about the multiplication algorithm used? Is it school grade, is it karatsuba? FFT?

As per firejuggler's suggestion, we tried to run this toy in an android emulator - our kingdom is Cortex M, but we have some android emus laying around, as we have colleagues who develop applications with them, for our customers. We wanted to see how the tests scale in time - if the application is bogus, like reading exponents from a list and saying prime/composite, and doing other things meantime, then the time-scaling would be also odd. We have some idea and could guess the multiplication algorithm used, if any tuning was done or it is just blind school-grade high-level stuff. And we consider this to be a good test, because it would not be easy for the guy to fake the process (which involves exponentiations and multiplications related not only to the size of the exponents, but also to the number of bits which are 1 involved). And our toys can also debug/disassemble code. But unluckily, we can not run it. Our emulator does not like it, and our colleagues who could help didn't come back from their NY holidays yet...

Our advice for now: avoid it.

P.S. Carlos, you seem to be the only one liking it? Can you detail why? Or was that a voting mistake? :razz:[/QUOTE]


Let's see colleague who is very good at insulting you without even being able to execute the application given your extensive experience. You say my program reads numbers from somewhere? Finally apprentice magician is very good insult without having much idea of ​​programming so it seems, make your an app for Movil that calculates numbers of Mersenne, here you are making comparisons with i7 and the cpu of a mobile. This program is for mobile and is the fastest on Android, the only one I've seen is a Google Play apk and it's very slow, this one is much faster and I'm talking about Android and mobile cpus, prime95 in AtoM n270 It is slower than this app in a GT-I9195. Greetings.

kladner 2019-01-05 01:37

[QUOTE=ewmayer;504960]Thanks - that glosses over the processor details, but that info is easily found - Here is [URL="https://www.anandtech.com/show/11540/samsung-galaxy-s8-exynos-versus-snapdragon"]AnandTech[/URL]:

That doesn't say precisely what "loosely based on" means - I'm hoping that at least the basic instruction set is identical.

By way of experimentation, we really need just one phone containing such an Arm-quad-based CPU and a suitable developer-tools interface of the kind app-writers use. In our case, said dev-tools could actually be just barebones gcc+libs.[/QUOTE]
Now I have to disavow my chip identification. I was going through apps, or things integrated with the OS, and came upon one with 'qualcomm' in its name. This shoots down the Samsung chip supposition, making it a Snapdragon octo, with [STRIKE]lower clocks than the Samsung[/STRIKE]. (WRONG. The Snapdragon has higher clocks.) I am not sure it matters to me, as I don't do heavy compute on the phone. I mostly use it to take pictures, text, and set reminders. I rarely talk on it.

The only regular web use is the Chicago Transit Authority Bus Tracker. It is generally quite accurate. It gives me a sense of whether I should jog to the corner, or just take my time. :smile:

thorken 2019-01-05 02:05

Calculate decimals of pi and e in Android. Spanish app
 
1 Attachment(s)
Here you have an installable apk of an application to calculate decimals of pi and e is fast but not the most and it is for Android who wants to try it

The app save the results in a txt file

[url]https://drive.google.com/file/d/1yLuUrMSDWfyrzUeeQph7kog0TzdJJDmP/view?usp=drivesdk[/url]

Batalov 2019-01-05 02:20

When I was 14, I calculated 100,000 digits of Pi with my own (obviously very simple) program. Now, let's see... this was 40 years ago and every 1.5 years the computational ability of modern hardware doubles, so that was done [B]on a computer that was 100,000,000 times slower than modern computers[/B]. ([URL="https://en.wikipedia.org/wiki/BESM-6"][SPOILER]BESM-6[/SPOILER][/URL])

Now if you compute 100,000 digits of Pi today, it will not even get you laid. And you computed, ... 2,000? :rolleyes:

Batalov 2019-01-05 02:37

[QUOTE=thorken;504979]Here you have an installable apk of an application to calculate decimals of pi and e is fast ...[/QUOTE]
...and of course, you should have first googled just for a minute or so.
You cannot easily compete with the likes of [URL="http://www.numberworld.org/y-cruncher/"]y-cruncher[/URL] and [URL="https://www.reddit.com/r/Android/comments/9xw9em/using_android_to_calculate_100000_digits_of_pi/"]Mini-Pi[/URL] for Pi and e.
[QUOTE]On the Pixel 3XL (snapdragon 845), computing 1,000,000 digits took around 5.8 seconds.[/QUOTE]You can definitely try - but for starters, you need to know what the state of the art is before beginning.

chalsall 2019-01-05 02:37

[QUOTE=Batalov;504980]Now if you compute 100,000 digits of Pi today, it will not even get you laid. And you computed, ... 2,000? :rolleyes:[/QUOTE]

LOL... Back in the day the girls who were will willing to trade compute for sex looked a bit like Romulans or Cardassians.

Now-a-days even a serious Amazon instance won't impress most human females....

Mysticial 2019-01-05 03:09

[QUOTE=Batalov;504981]...and of course, you should have first googled just for a minute or so.
You cannot easily compete with the likes of [URL="http://www.numberworld.org/y-cruncher/"]y-cruncher[/URL] and [URL="https://www.reddit.com/r/Android/comments/9xw9em/using_android_to_calculate_100000_digits_of_pi/"]Mini-Pi[/URL] for Pi and e.
You can definitely try - but for starters, you need to know what the state of the art is before beginning.[/QUOTE]

Calling Mini-Pi a "state of the art" is probably being a little too generous.

I quite literally slapped that thing together over a weekend with no effort. No real optimizations. Even GMP Pi is faster.

The reason why Mini-Pi gets any attention is probably because it's a single self-contained file.

thorken 2019-01-05 03:14

[QUOTE=Batalov;504981]...and of course, you should have first googled just for a minute or so.
You cannot easily compete with the likes of [URL="http://www.numberworld.org/y-cruncher/"]y-cruncher[/URL] and [URL="https://www.reddit.com/r/Android/comments/9xw9em/using_android_to_calculate_100000_digits_of_pi/"]Mini-Pi[/URL] for Pi and e.
You can definitely try - but for starters, you need to know what the state of the art is before beginning.[/QUOTE]

I know y cruncher and I repeat That this Apps are for mobile phones and With the app could calculate more decimals than 20000, this apks are for curiosity and for up temps of my Hands In winter,by the other hand the Apps run correctly and its fast not more fast than Apps use agm.

HERE a video of Mersenne app in Android and comparison With apk Lucas lehmer in Google play.

[URL="https://youtu.be/UR1jGwvUcQc"]https://youtu.be/UR1jGwvUcQc[/URL]

GP2 2019-01-05 03:23

[QUOTE=Batalov;504980]Now if you compute 100,000 digits of Pi today, it will not even get you laid.[/QUOTE]

Did that ever get anyone laid, at any point in human history?

Empirically [I like using that word], that sort of thing has the precisely opposite effect.

Or, uh... so I'm told.

kriesel 2019-01-05 16:13

[QUOTE=GP2;504990]Did that ever get anyone laid, at any point in human history?

Empirically [I like using that word], that sort of thing has the precisely opposite effect.

Or, uh... so I'm told.[/QUOTE]:rgerbicz::rofl:

kriesel 2019-01-05 16:50

[QUOTE=thorken;504975]Let's see colleague who is very good at insulting you without even being able to execute the application given your extensive experience. You say my program reads numbers from somewhere? Finally apprentice magician is very good insult without having much idea of ​​programming so it seems, make your an app for Movil that calculates numbers of Mersenne, here you are making comparisons with i7 and the cpu of a mobile. This program is for mobile and is the fastest on Android, the only one I've seen is a Google Play apk and it's very slow, this one is much faster and I'm talking about Android and mobile cpus, prime95 in AtoM n270 It is slower than this app in a GT-I9195. Greetings.[/QUOTE]
Sheesh, instead of steadfastly dodging it, just answer the question: "which algorithm of LL testing do you use ?" Or, what method is used to do the squaring? Here, I'll try to make it easy, multiple choice:

a) grammar school conventional long multiplication (using what "digit size") Twice the exponent, four times as many partial products to compute and sum, for each of twice as many iterations.

b) Karatsuba divide and conquer. Twice the exponent, [B]THREE[/B] times as many partial products, plus some additional overhead. (applied recursively, or not?)

c) Toom-Cook (for extra credit, which?)

d) modular multiplication (using which relative primes?)

e) fft (in which base?)

f) a mix depending on operand or exponent size (which, and at what breakpoints)

g) can't say, prepublication (what journal, and publication expected about when?)

h) super secret new faster than linear, proprietary, classified, NSA and KGB and Spain's National Intelligence Agency are after thorken for releasing it in executable form:shark:

i) Other (specify)

What programming language did you use?
Or put up some source code so we can look for ourselves.

(Are you by any chance acquainted with Billy Rubin?)

ewmayer 2019-01-05 20:10

Does anyone reading this have an Android dev-board with which to try an Mlucas build with? It would be best if it were one with an Arm CPU supporting the 128-bit SIMD instructions (a.k.a. Arm v8+), but even failing that one could try a generic C-code build using a scalar-doubles-based FFT.

(Of course the OP is welcome to try same, but at present OP seems to be busy fending off rude questions about his bigint multiply algorithm. ;)

science_man_88 2019-01-05 20:18

[QUOTE=thorken;504975]Let's see colleague who is very good at insulting you without even being able to execute the application given your extensive experience. You say my program reads numbers from somewhere? Finally apprentice magician is very good insult without having much idea of ​​programming so it seems, make your an app for Movil that calculates numbers of Mersenne, here you are making comparisons with i7 and the cpu of a mobile. This program is for mobile and is the fastest on Android, the only one I've seen is a Google Play apk and it's very slow, this one is much faster and I'm talking about Android and mobile cpus, prime95 in AtoM n270 It is slower than this app in a GT-I9195. Greetings.[/QUOTE]
They were just saying that's one possible way some program without source given could be both fast and accurate to what's already known. As shown a interpretted script can run faster. Also I got that lucas Lehmer app you talked about, when it takes 27-28 seconds for 12711 then yes many programs could run faster. that's why coders use asymptotics.

[QUOTE=ewmayer;505054]What exponents are those timings for, and on what compute hardware?[/QUOTE]

same exponents as OP, on the only working hardware I have at last check an android phone. I think it has a snapdragon processor, and PARIdroid was 2.11.0.1.5 edit2: seems my PARIdroid thinks they aren't prime though...

kriesel 2019-01-05 21:09

Ok, in fairness to thorken, in post one, he says he found the application on a phone, and makes no claims to having written it.

From his third screen shot:
[CODE]Analizando M9967 Con 3000.37 Digitos
Numeros Primos De Mersenne Encontrados 22
Tiempo Del Calculo 0h 2m 56s
Calculo Den Numeros Primos De Mersenne A 0.12 PrimosM/Seg
Indice De Velocidad 0.125 Puntos. MAX=100
Numeros Primos De Mersenne Calculados De 2^0-1 A 2^10000-1
Ultimo Numero Analizado Tenia 3000.37 Digitos
Archivo Creado En Tu Telefono Con Los Numeros Primos De Mersenne
((( ---- ||| ---- Fin Del Programa ----- ||| ----)))
Programado Por Marcos Ochoa Diez
...@gmail.com
http://youtube.com/...
MvIiIaX Corp. 2019
serranoltexx:/sdcard/download #[/CODE]Which Google translate converts (amusingly translating part of the name, Diez, to Ten) to
[CODE]Analyzing M9967 with 3000.37 Digits
Mersenne Primеs Found 22
Calculation Time 0h 2m 56s
I calculate Mersenne's Prime Numbers at 0.12 PrimosM / Sec
Speed Index 0.125 Points. MAX = 100
Mersenne Prime Numbers Calculated from 2 ^ 0-1 to 2 ^ 10000-1
Last Number Analyzed Tenia 3000.37 Digits
File Created On Your Phone With Mersenne Cousins Numbers
(((---- ||| ---- End Of Program ----- ||| ----)))
Programmed by Marcos Ochoa Ten
...@gmail.com
http://youtube.com/...
MvIiIaX Corp. 2019
serranoltexx: / sdcard / download #[/CODE]which, with a little human help, becomes
[CODE]Analyzing M9967 with 3000.37 Digits
Mersenne Primеs Found 22
Calculation Time 0h 2m 56s
I calculate Mersenne's Prime Numbers at 0.12 PrimosM / Sec
Speed Index 0.125 Points. MAX = 100
Mersenne Prime Numbers Calculated from 2^0 - 1 to 2^10000 - 1
Last Number Analyzed Had 3000.37 Digits
File Created On Your Phone With Mersenne Prime Numbers
(((---- ||| ---- End Of Program ----- ||| ----)))
Programmed by Marcos Ochoa Diez
...@gmail.com
http://youtube.com/...
MvIiIaX Corp. 2019
serranoltexx: / sdcard / download #[/CODE](Note though, that he did invite questions. The general rule of be careful what you ask for, particularly applies here on the forum.)

Searching online for Marcos Ochoa Diez yielded (and I by no means endorse any of it)
[URL]https://loteria-primitiva.uptodown.com/windows/descargar[/URL]
[URL]https://generador-de-numeros-primos.uptodown.com/windows[/URL]
[URL]https://www.amazon.com/Marcos-Ochoa-Diez-M8AX-Smoking/dp/B00RN9BYNS[/URL]
[URL]https://yepdownload.com/monitor-lunar-ip[/URL]
(Nothing on github or sourceforge in what I reviewed.)

pinhodecarlos 2019-01-05 21:23

“Tenia” translates into “had”.
“Mersenne cousin numbers”, please replace cousin with prime.

GP2 2019-01-05 23:12

[QUOTE=ewmayer;505069]So your simple Pari script running on Android is nearly 4x faster than [url=https://mersenneforum.org/showpost.php?p=504895&postcount=38]GP2's GMP-based LL tester[/url] running on an AWS Skylake core?[/QUOTE]

I didn't write it. It's adapted from [URL="http://rosettacode.org/wiki/Lucas-Lehmer_test#GMP"]RosettaCode[/URL], mostly by ripping out the preliminary TF stuff and making it print out a residue.

GMP isn't that complicated and I could have written a simple LL test myself, however this code incorporates an optimization for the modulo that probably wouldn't have occurred to me.

LaurV 2019-01-06 09:52

[QUOTE=pinhodecarlos;505060]“Tenia” translates into “had”.[/QUOTE]
You are wrong, I am sure it was about either [URL="https://en.wikipedia.org/wiki/Taenia_coli"]this[/URL] or better [URL="https://en.wikipedia.org/wiki/Taenia_(cestode)"]this[/URL]...

Batalov 2019-01-13 21:08

The [URL="https://mersenneforum.org/showthread.php?t=23995"]very fast running times thread is spun off[/URL], because these times cannot be comparable with any algorithm and confuses this discussion thread.


All times are UTC. The time now is 05:59.

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