Links to Precompiled GMP-ECM versions
 

This thread is for links to versions of GMP-ECM that have been compiled for various systems and discussions about their relative preformance. Please include information about target processors and operating systems and environments (i.e. Cygwin or MinGW or ?) and about the code and patch versions used.

GMP-ECM 6.0.1:

Athlon M4 Thunderbird 950Mhz. OS: Windows XP SP2. Compiled with MinGW: [URL=http://www.hoegge.dk/pub2/ecmathlon.rar]ecmathlon.rar[/URL]

P4 Northwood 2.4Ghz. OS: Windows XP SP2. Compiled with MinGW: [URL=http://www.hoegge.dk/pub2/ecmp4n.rar]ecmp4n.rar[/URL]

P4 Prescott 3.4Ghz. OS: Windows XP SP2. Compiled with MinGW: [URL=http://www.hoegge.dk/pub2/ecmp4p.rar]ecmp4p.rar[/URL]

Testing on Kevin G. Hare's c301 [URL=http://www.mersenneforum.org/showthread.php?t=3101]http://www.mersenneforum.org/showthread.php?t=3101[/URL]:

[U]P4 3.4Ghz:[/U]
ecm 3000000 < c301.txt
GMP-ECM 6.0.1 [powered by GMP 4.1.4] [ECM]
Input number is 3831565799519436303487742350308454794716675157894098584352121252263510024611805907320592374654433186020517108665467143471934035839395496243353321245760019611207664487665420776742726779780862993590544596916020496510980740067901995154639576852120198067468078357247366647828551141390739467161074462608561 (301 digits)
Using B1=3000000, B2=4016636513, polynomial Dickson(6), sigma=643218815
Step 1 took 161863ms
Step 2 took 53618ms

Same number on [U]Athlon 950Mhz:[/U]
Using B1=3000000, B2=4016636513, polynomial Dickson(6), sigma=3274898277
Step 1 took 376698ms
Step 2 took 134392ms

Same c301 on:
[U]P4 2.4Ghz[/U]
Using B1=3000000, B2=4016636513, polynomial Dickson(6), sigma=2830588821
Step 1 took 211290ms
Step 2 took 74878ms

Hi !

I have access to a Celeron M 360 (1.4Ghz, 1Mb fast L2 cache) and, whereas ggnfs is only 15 % slower with the celeron than with a 2Ghz OC Duron, P4 gmp-ecm pre-compiled binaries are 45-50% slower...
I know that Pentium M and Celeron M are P4 compatibles and I wonder if this difference is inherent to the program or if a pentium M gmp-ecm built would speed ecm a lot?? (ggnfs is Pentium M optimized)
BTW, can someone provide me pre-compiled pentium M win32 binaries or a link to such binaries (or simply an answer, to see if it's worth the time to install linux on the portable, just for compiling gmp and ecm myself with P.M optimizations...).

Thanks.
Regards.

Philippe.

Hi Philippe

You don't have to install Linux to compile an optimized ecm. If you're running Windoze you can try to follow [URL=http://www.geocities.com/omboohankvald/compileGMP.html]my guide[/URL] to compile it.

Good luck!

OH

[url=http://www.mystwalker.de/ECMnet/ecm-6.0.1.zip]Here[/url] are PentiumM-optimized binaries.

[QUOTE=OmbooHankvald]Hi Philippe

You don't have to install Linux to compile an optimized ecm. If you're running Windoze you can try to follow [URL=http://www.geocities.com/omboohankvald/compileGMP.html]my guide[/URL] to compile it.

Good luck!

OH[/QUOTE]

Hi OmbooHankvald!
I have formerly tried to compile gmp-ecm with cygwin (and I must say that compiling gmp with cygwin isn't so easy) and binaries were slightly less efficient than those compiled by others with more experience with gcc flags...That's why I was asking for binaries.
I do indeed appreciate your guides and have visited the page you show me a few days ago... Lacking a fast internet connexion, it is easier for me to download a binary than the whole mingw suite...
Thanks for helping me !

Hi Mystwalker!
Thanks a lot for the binaries!...I'll try them in a few minutes and, if interesting, post my results !

Best regards.
Philippe.

Anyone interested in compiling, for me, the GMP-ECM 6.0.1 for an AMD 64 windows machine?

Thanks in advance,

Carlos

Anyone?!?!?!

Carlos

Bump.

Are there 6.1 executables available for various configurations?

I've determined the 6.0.1 Windows executable I'm using on P4s and Athlon XPs fails when doing ECM with B1=260M, giving the message

GNU MP: Cannot allocate memory (size=562053120)

[QUOTE=wblipp]Bump.

Are there 6.1 executables available for various configurations?

I've determined the 6.0.1 Windows executable I'm using on P4s and Athlon XPs fails when doing ECM with B1=260M, giving the message

GNU MP: Cannot allocate memory (size=562053120)[/QUOTE]

You can reduce memory allocation by using various -k and -treefile options (I think), and [URL=http://www.geocities.com/fetofs/ecmzip.zip]here[/URL] is the binary I use for my P4 (compiled with GMP 4.2 and ECM 6.1).
I compiled it under WinXP/MinGW.

MacOS X, ECM 6.0.1 with GMP 4.2 [B]64-bit G5 only[/B]: [URL="http://www.greenbank.org/misc/ecm_6.0.1_gmp_4.2_MacOSX_64bit.zip"]ecm_6.0.1_gmp_4.2_MacOSX_64bit.zip[/URL]

gmp-ecm 6.1 w/ gmp 4.2, asm-redc disabled, ntt and tuned, available [url=http://www.mystwalker.de/ecm-6.1_dis_ntt_tune.zip]here[/url] :smile:

I found this combination to be the fastest in general, but not in every situation. Let's wait for gmp-ecm 6.1.1 to do a wide-range benchmarking...

[QUOTE=Mystwalker]gmp-ecm 6.1 w/ gmp 4.2, asm-redc disabled, ntt and tuned[/QUOTE]

Tuned for what?

[QUOTE=fetofs]Tuned for what?[/QUOTE]
I tuned the ecm-params.h parameters, using the "tune" command. I found that those were some % faster on a 500 digit candidate, but (IIRC) 1% slower on a 200 digit composite.
Most other compile combinations basically yielded equal run times per curve.

Notice that every other post describes what OS and platform their binaries have been compiled for. Notice what you post lacks. Now consider what was meant by:-

[QUOTE]Tuned for what?[/QUOTE]

[QUOTE=Mystwalker]I tuned the ecm-params.h parameters, using the "tune" command. I found that those were some % faster on a 500 digit candidate, but (IIRC) 1% slower on a 200 digit composite.
Most other compile combinations basically yielded equal run times per curve.[/QUOTE]

Hi !

I have noticed that it is very important to tune gmp-ecm 6.1 : look for the difference for stage 2
(most of the composites are ok with both versions but in this case, and some others, it makes a great difference) :

first compilation, without tuning :

GMP-ECM 6.1 [powered by GMP 4.2.1] [ECM]
Input number is 709891330215674922888729762564179876071621230485156504316706784486089//
628388958197669654729613064737223993515869067455247541707 (126 digits)
Using B1=3000000, B2=5706890290, polynomial Dickson(6), sigma=78228681
Step 1 took 78266ms
Step 2 took 101594ms ( ! )


recompiled, adjusting ecm-params.h :

GMP-ECM 6.1 [powered by GMP 4.2.1] [ECM]
Input number is 709891330215674922888729762564179876071621230485156504316706784486089//
628388958197669654729613064737223993515869067455247541707 (126 digits)
Using B1=3000000, B2=5706890290, polynomial Dickson(6), sigma=1397049393
Step 1 took 77828ms
Step 2 took 36078ms


This may explain some disappointing results with version 6.1...

Regards.
Philippe.

PS have you noticed that running many instances of tune.exe gives differents parameters? The first trial offers me good params but I am interested in increasing the accuracy of the test : how can I change the sources to obtain a better estimation?
Thanks

[QUOTE=Greenbank]Notice that every other post describes what OS and platform their binaries have been compiled for. Notice what you post lacks. Now consider what was meant by "Tuned for what?"[/QUOTE]

Ok, I just realized that it's not clear that my post #14 was a response to William's request in #11.
Hence, I explicitely state that it's a Windows binary. I don't know whether the optimal parameters differ across OS'es, though. If not, I've tuned it for every OS, one just needs to recompile for the resp. platform.

[QUOTE=Phil MjX]look for the difference for stage 2[/quote]

What CPU do you use? Pentium M and P4 only gained a few % in my tuning tests...

[quote]PS have you noticed that running many instances of tune.exe gives differents parameters?[/quote]

Yes, I also noticed this when I tested some dev-versions for David Newman. Some values deviate slightly, others from one extreme (e.g. 65535) to the other (e.g. 512).
Probably these values are not really critical or some background processes slightly modifiy the outcome of some parameters, which could result in other changed parameters.

But I don't have insight experience with this. I've only changed the values and did some benchmarking...

Thanks, fetofs and Mystwalker. Those solved the problem on the P4. Anybody have a version for Athlon XP?

William

Since something happened that "broke" OmbooHankvald's tutorial, I'd like to request an ecm.exe file that will work well on a 2.8GHz Pentium-D with 1MB cache for each core. I don't think they're shared, but I may be wrong. I've been having some really, REALLY bad luck with my new computer(I think the d: drive is causing most, if not all, of the trouble.), so I don't have a lot of stuff at the moment, like cpu-z.

[QUOTE=jasong]I'd like to request an ecm.exe file that will work well on a 2.8GHz Pentium-D with 1MB cache for each core.[/QUOTE]
Try the links in messages #12 and #14 in this thread.

Following GMP-ECM 6.1 binaries were compiled on Windows XP with Msys and MinGW (not the latest versions, does it matter?) and with GMP 4.2.1.
They were compiled with:
./configure --with-gmp=/usr/local
make
make check
make ecm-params; make
make install

Pentium4 Prescott 3.4 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm61-p4p.zip"]ecm61-p4p.zip[/URL]
Pentium4 Northwood 2.4 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm61-p4n.zip"]ecm61-p4n.zip[/URL]
Pentium4 Willamette 1.50 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm61-p4w.zip"]ecm61-p4w.zip[/URL]
Athlon XP 2200+ 1.80Ghz: [URL="http://www.hoegge.dk/mersenne/ecm61-athlon.zip"]ecm61-athlon.zip[/URL]

For those of you who won't, or can't, compile on their own:

I've discovered in my own experience that downloading an ecm.exe file made by someone else, as long as it's compatible, has improved by speed from 10.5 hours to 4.5 hours when comparing the ElevenSmooth client to one I downloaded in this thread.

GMP-ECM 6.1.1 with GMP 4.2.1 compiled on Windows XP with Msys+MinGW with:
./configure --with-gmp=/usr/local
make
make check
make ecm-params; make
make install

Pentium4 Prescott 3.4 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm611-p4p.zip"]ecm611-p4p.zip[/URL]
Pentium4 Northwood 2.4 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm611-p4n.zip"]ecm611-p4n.zip[/URL] (coming august 2nd)
Pentium4 Willamette 1.50 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm611-p4w.zip"]ecm611-p4w.zip[/URL] (coming august 7th-ish)
Pentium3 Celeron 1.10 Ghz (P6C Coppermine128): [URL="http://www.hoegge.dk/mersenne/ecm611-p3c.zip"]ecm611-p3c.zip[/URL]
Athlon XP 2200+ 1.80Ghz: [URL="http://www.hoegge.dk/mersenne/ecm611-athlon.zip"]ecm611-athlon.zip[/URL]

Thanks, ATH!

Did you already test on which architectures "--enable-asm-redc" gave you advantages?

No. I can try, is it enabled or disabled by default?

GMP-ECM 6.1.1 with GMP 4.2.1 compiled on Windows XP with Msys+MinGW. The asmredc files is compiled with --enable-asm-redc, check txt-files for speedtest: asm-redc seems slightly faster up to 190-200 digit numbers. No asm-redc on Pentium3 since no SSE2.

Pentium4 Prescott 3.4 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm611-p4p.zip"]ecm611-p4p.zip[/URL] [URL="http://www.hoegge.dk/mersenne/ecm611asmredc-p4p.zip"]ecm611asmredc-p4p.zip[/URL] [URL="http://www.hoegge.dk/mersenne/ecm611-p4p.txt"]ecm611-p4p.txt[/URL]
Pentium4 Northwood 2.4 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm611-p4n.zip"]ecm611-p4n.zip[/URL] [URL="http://www.hoegge.dk/mersenne/ecm611asmredc-p4n.zip"]ecm611asmredc-p4n.zip[/URL] [URL="http://www.hoegge.dk/mersenne/ecm611-p4n.txt"]ecm611-p4n.txt[/URL] (coming august 2nd)
Pentium4 Willamette 1.50 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm611-p4w.zip"]ecm611-p4w.zip[/URL] [URL="http://www.hoegge.dk/mersenne/ecm611asmredc-p4w.zip"]ecm611asmredc-p4w.zip[/URL] [URL="http://www.hoegge.dk/mersenne/ecm611-p4w.txt"]ecm611-p4w.txt[/URL] (coming august 7th-ish)
Pentium3 Celeron 1.10 Ghz (P6C Coppermine128): [URL="http://www.hoegge.dk/mersenne/ecm611-p3c.zip"]ecm611-p3c.zip[/URL] [URL="http://www.hoegge.dk/mersenne/ecm611-p3c.txt"]ecm611-p3c.txt[/URL]
Athlon XP 2200+ 1.80Ghz: [URL="http://www.hoegge.dk/mersenne/ecm611-athlon.zip"]ecm611-athlon.zip[/URL] [URL="http://www.hoegge.dk/mersenne/ecm611asmredc-athlon.zip"]ecm611asmredc-athlon.zip[/URL] [URL="http://www.hoegge.dk/mersenne/ecm611-athlon.txt"]ecm611-athlon.txt[/URL]

Have you got a precompiled version for Linux, too? I'm using a P4 and I'm not keen on installing GMP-ECM (and the GMP libraries)...

No sorry, I don't have Linux installed.

I have Linux on a couple of architectures, but sadly, no P4 among them. Only P3, Athlon, Opteron.

Alex

Following the instructions in the INSTALL and the README file I compiled a linux version for Athlon64, you can get it [url=http://www.informatik.hu-berlin.de/~schoenbe/ecm-6.1.1.tar.gz]here[/url].

System: Linux (64bit), Athlon64 (gmp 4.1.3, ecm 6.1.1)

I took the c301 from Kevin Hare for a quick test on an Athlon64 3400+ (@2.4GHz), program output:
[code]
ecm 3e6 < in.txt >> out.txt
GMP-ECM 6.1.1 [powered by GMP 4.1.3] [ECM]
Input number is 3831565799519436303487742350308454794716675157894098584352121252263510024611805907320592374654433186020517//1086654671434719340358393954962433533212457600196112076644876654207767427267797808629935905445969160204965//10980740067901995154639576852120198067468078357247366647828551141390739467161074462608561 (301 digits)
Using B1=3000000, B2=5706890290, polynomial Dickson(6), sigma=4115968196
Step 1 took 91619ms
Step 2 took 24242ms
[/code]
In comparison to my P4 it's really a nice speed (forget all those P4 processors...)

[QUOTE=Yamato;86091]Following the instructions in the INSTALL and the README file I compiled a linux version for Athlon64, you can get it [url=http://www.informatik.hu-berlin.de/~schoenbe/ecm-6.1.1.tar.gz]here[/url].

System: Linux (64bit), Athlon64 (gmp 4.1.3, ecm 6.1.1)

I took the c301 from Kevin Hare for a quick test on an Athlon64 3400+ (@2.4GHz), program output:
[code]
ecm 3e6 < in.txt >> out.txt
GMP-ECM 6.1.1 [powered by GMP 4.1.3] [ECM]
Input number is 3831565799519436303487742350308454794716675157894098584352121252263510024611805907320592374654433186020517//1086654671434719340358393954962433533212457600196112076644876654207767427267797808629935905445969160204965//10980740067901995154639576852120198067468078357247366647828551141390739467161074462608561 (301 digits)
Using B1=3000000, B2=5706890290, polynomial Dickson(6), sigma=4115968196
Step 1 took 91619ms
Step 2 took 24242ms
[/code]
In comparison to my P4 it's really a nice speed (forget all those P4 processors...)[/QUOTE]Interesting.

I ran your binary on my 3500+ with this result:
[code]pcl@imhotep ~/ecm-6.1.1 $ ./ecm 3000000
GMP-ECM 6.1.1 [powered by GMP 4.1.4] [ECM]
3831565799519436303487742350308454794716675157894098584352121252263510024611805907320592374654433186020517108665467143471934035839395496243353321245760019611207664487665420776742726779780862993590544596916020496510980740067901995154639576852120198067468078357247366647828551141390739467161074462608561
Input number is 3831565799519436303487742350308454794716675157894098584352121252263510024611805907320592374654433186020517108665467143471934035839395496243353321245760019611207664487665420776742726779780862993590544596916020496510980740067901995154639576852120198067468078357247366647828551141390739467161074462608561 (301 digits)
Using B1=3000000, B2=5706890290, polynomial Dickson(6), sigma=2206158585
Step 1 took 48445ms
Step 2 took 18764ms
[/code]
This is approximately twice the speed of your run for stage 1 and noticeably faster in stage 2. The major difference, as I see it, is that I've GMP 4.1.4 installed. The clock speed of my system is [i]slower[/i] than yours (2188.896 MHz) according to /proc/cpuinfo.

Performing the same test on the ecm 6.0.1 I have installed gives this result:
[code]pcl@imhotep ~/ecm-6.1.1 $ /usr/local/bin/ecm 3000000
GMP-ECM 6.0.1 [powered by GMP 4.1.4] [ECM]
3831565799519436303487742350308454794716675157894098584352121252263510024611805907320592374654433186020517108665467143471934035839395496243353321245760019611207664487665420776742726779780862993590544596916020496510980740067901995154639576852120198067468078357247366647828551141390739467161074462608561
Input number is 3831565799519436303487742350308454794716675157894098584352121252263510024611805907320592374654433186020517108665467143471934035839395496243353321245760019611207664487665420776742726779780862993590544596916020496510980740067901995154639576852120198067468078357247366647828551141390739467161074462608561 (301 digits)
Using B1=3000000, B2=4016636513, polynomial Dickson(6), sigma=1196414020
Step 1 took 50620ms
Step 2 took 26185ms
[/code]

So 6.1.1 is marginally faster in stage 1 and substantially faster in stage 2 despite having a markedly higher B2 value. Both programs run much faster on my machine, despite it having a lower clock speed.

Quite remarkable. I'm wondering whether your GMP has the x64 assembler code incorporated.


Paul

GMP is statically linked into GMP-ECM so which version of GMP is installed on the system it runs on should not matter, only which version was installed on the system it was compiled on. The different speeds are rather surprising.

Alex

Note: Yamato's post about build problems moved to [url]http://www.mersenneforum.org/showthread.php?t=6186[/url]

[QUOTE=akruppa;85530]I have Linux on a couple of architectures, but sadly, no P4 among them. Only P3, Athlon, Opteron.

Alex[/QUOTE]

Hi Alex, is that Athlon 64?

I have access to a few Athlon, Athlon XP, and Opteron. Opteron uses the AMD64 architecture just like Athlon 64, so I can make binaries for Athlon 64 on the Opterons.

Alex

I've rebuild the binary with gmp 4.2.1 - it is quite faster now.

System: Linux (64bit), Athlon64 (gmp 4.2.1, ecm 6.1.1)
File: [url=http://www.informatik.hu-berlin.de/~schoenbe/ecm-6.1.1.tar.gz]ecm-6.1.1.tar.gz[/url]

Hare's C301 on Athlon64 3400+:

[code]
ecm 3e6 < in.txt >> out.txt
GMP-ECM 6.1.1 [powered by GMP 4.2.1] [ECM]
Input number is 3831565799519436303487742350308454794716675157894098584352121252263510024611805907320592374654433186020517108665467143471934035839395496243353321245760019611207664487665420776742726779780862993590544596916020496510980740067901995154639576852120198067468078357247366647828551141390739467161074462608561 (301 digits)
Using B1=3000000, B2=5706890290, polynomial Dickson(6), sigma=2465834108
Step 1 took 52143ms
Step 2 took 20238ms
[/code]

Compare these times with those of the P4 3.4GHz (see ATH's posting). That seems to me a stunning improvement within 15 months.

[quote]
The clock speed of my system is slower than yours (2188.896 MHz) according to /proc/cpuinfo.
[/quote]

IIRC the 3500+ has 1024K cache, the 3400+ has only 512K, it might possibly have a significant effect.

[QUOTE=akruppa;86108]GMP is statically linked into GMP-ECM so which version of GMP is installed on the system it runs on should not matter, only which version was installed on the system it was compiled on. The different speeds are rather surprising.[/QUOTE]
Perhaps so.

My experience suggests otherwise. Take a look at the output I quoted. You will note that my machine reported 4.1.4 but that Yamato's reported 4.1.3

Unless he rebuilt it on 4.1.4 after taking his screenshot ...


Paul

Is the readme file hosted anywhere for someone playing around with pre-compiled versions?

Not as far as I know. It's only in the .tar.gz file.

Alex

Is there a way to unzip it (under Windows) without paying out $40 for some unzipping software?

[URL="http://www.download.com/WinRAR/3000-2250_4-10583567.html?tag=lst-0-1"]WinRAR[/URL] is free for 40-days, and it works.

but here is readme, news, install from GMP-ECM 6.1.1, I added .txt to them because I'm a windows freak:

[URL="http://www.hoegge.dk/mersenne/README.txt"]README.txt[/URL]
[URL="http://www.hoegge.dk/mersenne/INSTALL.txt"]INSTALL.txt[/URL]
[URL="http://www.hoegge.dk/mersenne/NEWZ.txt"]NEWZ.txt[/URL]

Thanks, I just found the version 5 readme at ElevenSmooth.

[QUOTE=philmoore;87962]Is there a way to unzip it (under Windows) without paying out $40 for some unzipping software?[/QUOTE]

You can use [url=http://www.7-zip.org/]7-Zip[/url].

[QUOTE=ATH;87966][URL="http://www.download.com/WinRAR/3000-2250_4-10583567.html?tag=lst-0-1"]WinRAR[/URL] is free for 40-days, and it works.

but here is readme, news, install from GMP-ECM 6.1.1, I added .txt to them because I'm a windows freak:

[URL="http://www.hoegge.dk/mersenne/README.txt"]README.txt[/URL]
[URL="http://www.hoegge.dk/mersenne/INSTALL.txt"]INSTALL.txt[/URL]
[URL="http://www.hoegge.dk/mersenne/NEWZ.txt"]NEWZ.txt[/URL][/QUOTE]

You have got a typo in the last link (newZ instead of newS) - here is the correct link:

[URL="http://www.hoegge.dk/mersenne/NEWS.txt"]NEWS.txt[/URL]

Thanks, I had found the version 6.0.1 readme at
[url]http://web.mit.edu/afs/sipb/project/pari-gp/gmp-ecm/src/ecm-6.0.1/[/url]
but it is nice to have the latest version.

[QUOTE]You have got a typo in the last link (newZ instead of newS) - here is the correct link:[/QUOTE]

LOL thanks, thats because I daily check a danish webpage called [url]www.newz.dk[/url] , so I sometimes spell news with newz now heh.

Hello,

Has someone of you a precompiled version for Linux on Core 2 Duo?

I have access to several Core 2 machines (but not the permission to install software) and want to test them.

GMP-ECM 6.1.2 with GMP 4.2.1 compiled on Windows XP with Msys+MinGW with:
./configure --with-gmp=/usr/local
make
make check
make ecm-params; make
make install

Those files called -asmredc are compiled with --enable-asm-redc switch, which in GMP-ECM 6.1.1 was faster up to 190-200 digit numbers.

Pentium 4 550 (Prescott) 3.4 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm612-p4p.zip"]ecm612-p4p.zip[/URL] [URL="http://www.hoegge.dk/mersenne/ecm612-p4p-asmredc.zip"]ecm612-p4p-asmredc.zip[/URL]
Pentium 4 (Northwood) 2.4 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm612-p4n.zip"]ecm612-p4n.zip[/URL] [URL="http://www.hoegge.dk/mersenne/ecm612-p4n-asmredc.zip"]ecm612-p4n-asmredc.zip[/URL]
Pentium 4 (Willamette) 1.50 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm612-p4w.zip"]ecm612-p4w.zip[/URL] [URL="http://www.hoegge.dk/mersenne/ecm612-p4w-asmredc.zip"]ecm612-p4w-asmredc.zip[/URL]
Pentium 3 (Coppermine) Celeron 1.10 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm612-p3c.zip"]ecm612-p3c.zip[/URL]
Athlon XP (Thoroughbred) 2200+ 1.80 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm612-athlon.zip"]ecm612-athlon.zip[/URL] [URL="http://www.hoegge.dk/mersenne/ecm612-athlon-asmredc.zip"]ecm612-athlon-asmredc.zip[/URL]

GMP-ECM 6.1.3 with GMP 4.2.1 compiled on Windows XP with Msys+MinGW with:
./configure --with-gmp=/usr/local
make
make check
make ecm-params; make
make install

Those files called -asmredc are compiled with --enable-asm-redc switch (Montgomery's REDC algorithm), which in GMP-ECM 6.1.1 was faster up to 190-200 digit numbers.

Pentium 4 550 (Prescott) 3.4 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm613-p4p.zip"]ecm613-p4p.zip[/URL] [URL="http://www.hoegge.dk/mersenne/ecm613-p4p-asmredc.zip"]ecm613-p4p-asmredc.zip[/URL]
Pentium 4 (Northwood) 2.4 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm613-p4n.zip"]ecm613-p4n.zip[/URL] [URL="http://www.hoegge.dk/mersenne/ecm613-p4n-asmredc.zip"]ecm613-p4n-asmredc.zip[/URL]
Pentium 4 (Willamette) 1.50 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm613-p4w.zip"]ecm613-p4w.zip[/URL] [URL="http://www.hoegge.dk/mersenne/ecm613-p4w-asmredc.zip"]ecm613-p4w-asmredc.zip[/URL]
Athlon XP (Thoroughbred) 2200+ 1.80 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm613-athlon.zip"]ecm613-athlon.zip[/URL] [URL="http://www.hoegge.dk/mersenne/ecm613-athlon-asmredc.zip"]ecm613-athlon-asmredc.zip[/URL]

Hello !

Thanks a lot for the binaries.
A question : have you tried to compile gmp-ecm 6.1.3 with the new gmp 4.2.2 library (actually a RC3) ? I think the release of the gmp version was the occasion to deliver the bug fix gmp-ecm version...
I am wondering if gmp 4.2.2 will speed up ecm computations...

any experiments ?

Best regards.

Philippe.

GMP-ECM 6.1.3 with GMP 4.2.1 compiled on Athlon64 (Linux). Both made with
./configure
make
make check
make ecm-params; make
make install

Linux (32 Bit): [url=http://www.informatik.hu-berlin.de/~schoenbe/ecm-6.1.3-32bit.tar.gz]ecm-6.1.3-32bit.tar.gz[/url]
Linux (64 Bit): [url=http://www.informatik.hu-berlin.de/~schoenbe/ecm-6.1.3-64bit.tar.gz]ecm-6.1.3-64bit.tar.gz[/url]

GMP-ECM 6.1.3 with [I][B]GMP 4.2.2[/B][/I] compiled on Windows XP with Msys+MinGW with:
./configure --with-gmp=/usr/local
make
make check
make ecm-params; make
make install

Those files called -asmredc are compiled with --enable-asm-redc switch (Montgomery's REDC algorithm), which in GMP-ECM 6.1.1 was faster up to 190-200 digit numbers.

Pentium 4 550 (Prescott) 3.4 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm613-gmp422-p4p.zip"]ecm613-gmp422-p4p.zip[/URL] [URL="http://www.hoegge.dk/mersenne/ecm613-gmp422-p4p-asmredc.zip"]ecm613-gmp422-p4p-asmredc.zip[/URL]
Pentium 4 (Northwood) 2.4 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm613-gmp422-p4n.zip"]ecm613-gmp422-p4n.zip[/URL] [URL="http://www.hoegge.dk/mersenne/ecm613-gmp422-p4n-asmredc.zip"]ecm613-gmp422-p4n-asmredc.zip[/URL]
Pentium 4 (Willamette) 1.50 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm613-gmp422-p4w.zip"]ecm613-gmp422-p4w.zip[/URL] [URL="http://www.hoegge.dk/mersenne/ecm613-gmp422-p4w-asmredc.zip"]ecm613-gmp422-p4w-asmredc.zip[/URL]
Athlon XP (Thoroughbred) 2200+ 1.80 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm613-gmp422-athlon.zip"]ecm613-gmp422-athlon.zip[/URL] [URL="http://www.hoegge.dk/mersenne/ecm613-gmp422-athlon-asmredc.zip"]ecm613-gmp422-athlon-asmredc.zip[/URL]

[QUOTE=Yamato;104471]Has someone of you a precompiled version for Linux on Core 2 Duo?[/QUOTE]

[url=http://www.geocities.com/g_w_reynolds/ecm/]Here[/url] is a x86-64 Linux binary optimised for Core 2.

[QUOTE=geoff;114882][url=http://www.geocities.com/g_w_reynolds/ecm/]Here[/url] is a x86-64 Linux binary optimised for Core 2.[/QUOTE]

Ouch. Core 2 duos are 64-bit; and I'm betting Prescott is 32-bit --- looks
like woodcrest is the 64-bit one (!?). Any chance of a windows 64-bit
binary for core 2 duo?

Next question. Is the above a candidate for comparison on a
linux dual xeon quad? Sounds like a single dual quad is getting
8 condor vm's, so I'm going to prefer something that stays on
its own core. I recently got a linux 32-bit xeon binary compiled
with gmp-4.2.2, but I'm certain that I haven't followed .001 of
Torbjorn's suggestions --- not sure what to make of the jmu
core2 assembly; and as I've reported elsewhere, Torbjorn's on
record with the view that xeon's are mostly suitable for use in
keeping rooms warm, rather than computing --- I'm reluctant to
ask. Any help would be appreciated. -Bruce

[QUOTE=bdodson;115724]Any chance of a windows 64-bit
binary for core 2 duo? [/QUOTE]

I can't get GMP to compile for x86-64 Windows, the problem seems to be that long int is 32-bits wide and GMP expects it to be 64 bits (as it is on every other x86-64 system).

[QUOTE]Next question. Is the above a candidate for comparison on a
linux dual xeon quad?[/QUOTE]
No idea, sorry.

[QUOTE=geoff;116850]I can't get GMP to compile for x86-64 Windows, the problem seems to be that long int is 32-bits wide and GMP expects it to be 64 bits (as it is on every other x86-64 system).

No idea, sorry.[/QUOTE]

Thanks for saving me a trek over to the library; I was considering
heading over to see whether I could get ATH's min-what's-it to give
a 64-bit windows binary for the core2's, but that sounds not a likely
prospect.
At the time I wrote, I was probably more concerned about the
even newer dual xeon quads/linux, but gmp-ecm seemed fairly happy with
them, and gave a 64-bit binary that runs circles around the 32-bit
compile. The Opterons 64-bit binary is quicker, but we got 3-times-40
clustered Opterons; while the new cluster has 8-times-40 cores, as well
as 2Gb/core, instead of 1Gb. So far as I've been able to tell, the
scheduling software isn't quite up to keeping all of the cores busy yet,
but it's still very early (and no one else is running yet!). -Bruce

[QUOTE=bdodson;116859] ... I was probably more concerned about the
even newer dual xeon quads/linux, but gmp-ecm seemed fairly happy with
them, and gave a 64-bit binary that runs circles around the 32-bit
compile. ... the new cluster has 8-times-40 cores, as well
as 2Gb/core ... So far as I've been able to tell, the scheduling software
isn't quite up to keeping all of the cores busy yet, but it's still very early ...
[/QUOTE]

Ooops, spoke too soon, before giving our sys staff time to respond
to yesterday's am email. This morning's running core count is
57 AMD-64's/Opterons, to 301 cores from the quads; so condor is
definitely potentially up to the scheduling (hoping it stays up for a while ...).

Strangely, the Opterons seem to be responding to challenge, with a
second recent factor (after some months without any), a base-2, no
less,

p54 = 310379797028432711386895309334277387365875751630693121

from 2, 832+, c228 --> c175. -Bruce

[QUOTE=bdodson;116859]Thanks for saving me a trek over to the library; I was considering heading over to see whether I could get ATH's min-what's-it to give a 64-bit windows binary for the core2's, but that sounds not a likely prospect.
[/QUOTE]

I hear that it is possible to compile GMP for x86-64 with MSVC, and it could be cross-compiling from Linux that is causing the problems for me with GCC. There is a Windows GCC binary package that can produce x86-64 Windows executables at [url]http://sourceforge.net/projects/mingw-w64/[/url] if anyone wants to have a go from Windows.

Comparing a 32-bit P4 optimised executable against the 64-bit Core 2 optimised executable, both running on a Core 2 Duo E6750, the 64-bit executable was typically about 90% faster in stage 1 and about 70% faster in stage 2.

GMP-ECM 6.1.3 with [I][B]GMP 4.2.2[/B][/I] compiled on Windows XP with Msys+MinGW with:

Core2 Duo (Conroe) E6750 2.66 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm613-gmp422-c2d.zip"]ecm613-gmp422-c2d.zip[/URL] --enable-asm-redc: [URL="http://www.hoegge.dk/mersenne/ecm613-gmp422-c2d-asmredc.zip"]ecm613-gmp422-c2d-asmredc.zip[/URL]

Could someone please post a Windows Opteron ecm? Windows XP, if it matters.

could somone post a ecm6.2 rc1 binary for athlonxp

[B][I]GMP-ECM 6.2rc1[/I][/B] with GMP 4.2.2 compiled on Windows XP/Vista(32bit not 64bit) with Msys+MinGW with:
./configure --with-gmp=/usr/local
make
make check
make ecm-params; make
make install

Core2Duo (Conroe) E6750 2.66 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm62rc1-c2d.zip"]ecm62rc1-c2d.zip[/URL](32bit) --enable-asm-redc: [URL="http://www.hoegge.dk/mersenne/ecm62rc1-c2d-asmredc.zip"]ecm62rc1-c2d-asmredc.zip[/URL](32bit)
Mobile Core2Duo T7300 (Merom) 2.00 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm62rc1-mobc2d.zip"]ecm62rc1-mobc2d.zip[/URL](32bit) --enable-asm-redc: [URL="http://www.hoegge.dk/mersenne/ecm62rc1-mobc2d-asmredc.zip"]ecm62rc1-mobc2d-asmredc.zip[/URL](32bit)
Pentium 4 550 (Prescott) 3.4 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm62rc1-p4p.zip"]ecm62rc1-p4p.zip[/URL] --enable-asm-redc: [URL="http://www.hoegge.dk/mersenne/ecm62rc1-p4p-asmredc.zip"]ecm62rc1-p4p-asmredc.zip[/URL]
Pentium 4 (Northwood) 2.4 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm62rc1-p4n.zip"]ecm62rc1-p4n.zip[/URL] --enable-asm-redc:[URL="http://www.hoegge.dk/mersenne/ecm62rc1-p4n-asmredc.zip"]ecm62rc1-p4n-asmredc.zip[/URL]
Athlon XP (Thoroughbred) 2200+ 1.80 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm62rc1-athlon.zip"]ecm62rc1-athlon.zip[/URL] --enable-asm-redc: [URL="http://www.hoegge.dk/mersenne/ecm62rc1-athlon-asmredc.zip"]ecm62rc1-athlon-asmredc.zip[/URL]

Were there any problems making those binaries?

Btw, I should have mentioned this before (or enabled asserts by default for the rc): can you please test the release candidates with "--enable-assert" ? It adds a lot of tests to the code that will hopefully help discovery bugs.

Alex

I pulled the svn version again this afternoon and compiled a Win64 version using Brian's VC project files (which works great). I tried a couple ways of putting "--enable-assert" in there but none of them worked. Does anyone know if it's possible to use that compile option in VC? I can post the Win64 binary anyway, but if someone knows how to enable that option, it may be more helpful for testing.

[quote=ATH;132528][B][I]GMP-ECM 6.2rc1[/I][/B] with GMP 4.2.2 compiled on Windows XP/Vista(32bit not 64bit) with Msys+MinGW with:
./configure --with-gmp=/usr/local
make
make check
make ecm-params; make
make install

Core2Duo (Conroe) E6750 2.66 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm62rc1-c2d.zip"]ecm62rc1-c2d.zip[/URL](32bit) --enable-asm-redc: [URL="http://www.hoegge.dk/mersenne/ecm62rc1-c2d-asmredc.zip"]ecm62rc1-c2d-asmredc.zip[/URL](32bit)
Mobile Core2Duo T7300 (Merom) 2.00 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm62rc1-mobc2d.zip"]ecm62rc1-mobc2d.zip[/URL](32bit) --enable-asm-redc: [URL="http://www.hoegge.dk/mersenne/ecm62rc1-mobc2d-asmredc.zip"]ecm62rc1-mobc2d-asmredc.zip[/URL](32bit)
Pentium 4 550 (Prescott) 3.4 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm62rc1-p4p.zip"]ecm62rc1-p4p.zip[/URL] --enable-asm-redc: [URL="http://www.hoegge.dk/mersenne/ecm62rc1-p4p-asmredc.zip"]ecm62rc1-p4p-asmredc.zip[/URL]
Pentium 4 (Northwood) 2.4 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm62rc1-p4n.zip"]ecm62rc1-p4n.zip[/URL] --enable-asm-redc:[URL="http://www.hoegge.dk/mersenne/ecm62rc1-p4n-asmredc.zip"]ecm62rc1-p4n-asmredc.zip[/URL]
Athlon XP (Thoroughbred) 2200+ 1.80 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm62rc1-athlon.zip"]ecm62rc1-athlon.zip[/URL] --enable-asm-redc: [URL="http://www.hoegge.dk/mersenne/ecm62rc1-athlon-asmredc.zip"]ecm62rc1-athlon-asmredc.zip[/URL][/quote]
thanks

[QUOTE=jbristow;132537]I pulled the svn version again this afternoon and compiled a Win64 version using Brian's VC project files (which works great). I tried a couple ways of putting "--enable-assert" in there but none of them worked. Does anyone know if it's possible to use that compile option in VC? I can post the Win64 binary anyway, but if someone knows how to enable that option, it may be more helpful for testing.[/QUOTE]

Brian's project comes with a config.h and everything, there is no need (and no use) in running the configure script again. To change parameters you'd normally give as options to configure, you need to edit the config.h file directly. The
#undef WANT_ASSERT
line should be changed to
#define WANT_ASSERT 1
to enable assertions.

Alex

Thanks Alex.

It looks like the zipped Win64 binaries are just slightly too large to post here (260kb vs. a 244kb limit). They were compiled on a Core 2 Quad Q6600. If anyone wants one, let me know.

[QUOTE=jbristow;132557]Thanks Alex.

It looks like the zipped Win64 binaries are just slightly too large to post here (260kb vs. a 244kb limit). They were compiled on a Core 2 Quad Q6600. If anyone wants one, let me know.[/QUOTE]

Can you compress the files with bzip2? I experienced that this is more effective (i.e. smaller files) than zip in most cases.

[QUOTE=akruppa;132530]Were there any problems making those binaries?

Btw, I should have mentioned this before (or enabled asserts by default for the rc): can you please test the release candidates with "--enable-assert" ? It adds a lot of tests to the code that will hopefully help discovery bugs.

Alex[/QUOTE]

Do I use "--enable-assert" as standard? or is this for a special test binary?

[quote=Andi47;132563]Can you compress the files with bzip2? I experienced that this is more effective (i.e. smaller files) than zip in most cases.[/quote]

Yeah, I tried bzip2 and it turned out about the same.

[QUOTE=ATH;132564]Do I use "--enable-assert" as standard? or is this for a special test binary?[/QUOTE]

Assertions make the code a little slower (by a few percent). If you think that GMP-ECM is bug-free enough for your liking, you can leave them off for a slight speedup, but for testing I recommend to switch them on.

Alex

[QUOTE=akruppa;132569]Assertions make the code a little slower (by a few percent). If you think that GMP-ECM is bug-free enough for your liking, you can leave them off for a slight speedup, but for testing I recommend to switch them on.

Alex[/QUOTE]

[B][I]GMP-ECM 6.2rc1[/I][/B] with GMP 4.2.2 with --enable-assert:

Core2Duo (Conroe) E6750 2.66 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm62rc1-c2d-assert.zip"]ecm62rc1-c2d-assert.zip[/URL](32bit)
Mobile Core2Duo T7300 (Merom) 2.00 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm62rc1-mobc2d-assert.zip"]ecm62rc1-mobc2d-assert.zip[/URL](32bit)
Pentium 4 550 (Prescott) 3.4 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm62rc1-p4p-assert.zip"]ecm62rc1-p4p-assert.zip[/URL]
Pentium 4 (Northwood) 2.4 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm62rc1-p4n-assert.zip"]ecm62rc1-p4n-assert.zip[/URL]
Athlon XP (Thoroughbred) 2200+ 1.80 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm62rc1-athlon-assert.zip"]ecm62rc1-athlon-assert.zip[/URL]

GMP-ECM 6.2[B][I]rc2[/I][/B] with GMP 4.2.2 compiled on Windows XP/Vista(32bit not 64bit) with Msys+MinGW with:
./configure --with-gmp=/usr/local
make
make check
make ecm-params; make
make install

Core2Duo (Conroe) E6750 2.66 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm62rc2-c2d.zip"]ecm62rc2-c2d.zip[/URL](32bit) --enable-asm-redc: [URL="http://www.hoegge.dk/mersenne/ecm62rc2-c2d-asmredc.zip"]ecm62rc2-c2d-asmredc.zip[/URL](32bit)
Mobile Core2Duo T7300 (Merom) 2.00 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm62rc2-mobc2d.zip"]ecm62rc2-mobc2d.zip[/URL](32bit) --enable-asm-redc: [URL="http://www.hoegge.dk/mersenne/ecm62rc2-mobc2d-asmredc.zip"]ecm62rc2-mobc2d-asmredc.zip[/URL](32bit)
Pentium 4 550 (Prescott) 3.4 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm62rc2-p4p.zip"]ecm62rc2-p4p.zip[/URL] --enable-asm-redc: [URL="http://www.hoegge.dk/mersenne/ecm62rc2-p4p-asmredc.zip"]ecm62rc2-p4p-asmredc.zip[/URL]
Pentium 4 (Northwood) 2.4 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm62rc2-p4n.zip"]ecm62rc2-p4n.zip[/URL] --enable-asm-redc:[URL="http://www.hoegge.dk/mersenne/ecm62rc2-p4n-asmredc.zip"]ecm62rc2-p4n-asmredc.zip[/URL]
Athlon XP (Thoroughbred) 2200+ 1.80 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm62rc2-athlon.zip"]ecm62rc2-athlon.zip[/URL] --enable-asm-redc: [URL="http://www.hoegge.dk/mersenne/ecm62rc2-athlon-asmredc.zip"]ecm62rc2-athlon-asmredc.zip[/URL]

GMP-ECM [B][I]6.2[/I][/B] with GMP 4.2.2 compiled on Windows XP/Vista(32bit not 64bit) with Msys+MinGW:

Core2Duo (Conroe) E6750 2.66 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm62-c2d.zip"]ecm62-c2d.zip[/URL](32bit) --enable-asm-redc: [URL="http://www.hoegge.dk/mersenne/ecm62-c2d-asmredc.zip"]ecm62-c2d-asmredc.zip[/URL](32bit)
Mobile Core2Duo T7300 (Merom) 2.00 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm62-mobc2d.zip"]ecm62-mobc2d.zip[/URL](32bit) --enable-asm-redc: [URL="http://www.hoegge.dk/mersenne/ecm62-mobc2d-asmredc.zip"]ecm62-mobc2d-asmredc.zip[/URL](32bit)
Pentium 4 550 (Prescott) 3.4 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm62-p4p.zip"]ecm62-p4p.zip[/URL] --enable-asm-redc: [URL="http://www.hoegge.dk/mersenne/ecm62-p4p-asmredc.zip"]ecm62-p4p-asmredc.zip[/URL]
Pentium 4 (Northwood) 2.4 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm62-p4n.zip"]ecm62-p4n.zip[/URL] --enable-asm-redc:[URL="http://www.hoegge.dk/mersenne/ecm62-p4n-asmredc.zip"]ecm62-p4n-asmredc.zip[/URL]
Athlon XP (Thoroughbred) 2200+ 1.80 Ghz: [URL="http://www.hoegge.dk/mersenne/ecm62-athlon.zip"]ecm62-athlon.zip[/URL] --enable-asm-redc: [URL="http://www.hoegge.dk/mersenne/ecm62-athlon-asmredc.zip"]ecm62-athlon-asmredc.zip[/URL]
(athlon+p4n tomorrow and core2duo tuesday)

[quote=jbristow;132537]I pulled the svn version again this afternoon and compiled a Win64 version using Brian's VC project files (which works great). I tried a couple ways of putting "--enable-assert" in there but none of them worked. Does anyone know if it's possible to use that compile option in VC? I can post the Win64 binary anyway, but if someone knows how to enable that option, it may be more helpful for testing.[/quote]

Hi to All,

I have not posted often here in the past but now that I am maintaining a Microsoft Visual Studio 2008 build of GMP-ECM, I thought that I should mention that I am willing to offer advice to anyone who is having problems building GMP with the native Windows tools.

I much appreciate the help that Alex and Paul have provided in building a robust 64-bit Windows build of GMP-ECM. This is harder on Windows than it might seem because long types are 32-bits on Windows and this makes it necessary to use long long types to obtain 64-bit limbs. It is a tribute to both the GMP team and the LORIA/INRIA team that this all now works.

I also thank Alex and Paul for motivating me to find out why my Windows port of GMP has previosuly had relatively poor perfromance on division. I have now sorted this out and I get a gmpbench score of 8950 on a 2.4GHz Athlon X2.

Turning to GMP-ECM itself, it is worth mentioning that my build files don't turn on OpenMP support. But this can be used with the Microsoft compiler so if you have a multi core processor it might be worth trying this out (the setting is in the 'C/C++|language' property page - it needs to be set for both libecm and ecm).

with regards to all

Brian Gladman

could someone post a 64-bit linux binary of 6.2.1 please

[QUOTE=henryzz;136378]could someone post a 64-bit linux binary of 6.2.1 please[/QUOTE]

I have put a 64-bit Linux executable optimised for Core 2 at [url]http://www.geocities.com/g_w_reynolds/ecm/[/url]

[QUOTE=geoff;137078]I have put a 64-bit Linux executable optimised for Core 2 at [url]http://www.geocities.com/g_w_reynolds/ecm/[/url][/QUOTE]

Does it include GWNUM library? I alerady compiled the base version for Linux 64. :smile:

Luigi

[QUOTE=ET_;137090]Does it include GWNUM library? I alerady compiled the base version for Linux 64. :smile:

Luigi[/QUOTE]

It is statically linked against libgmp.a version 4.2.2 with JWM's Core 2 patches applied, so it won't use the installed libgmp.so. It still uses the other installed system libraries.

[QUOTE=geoff;137078]I have put a 64-bit Linux executable optimised for Core 2 at [url]http://www.geocities.com/g_w_reynolds/ecm/[/url][/QUOTE]

What parameters did you use to compile it?

I'm trying to use the Msys and the MinGW 64 bit to compile it on Windows x64, but I can't get it to work yet.


I tried to get the complete svn package with TortoiseSVN, but it ended in an error. I'm new to this Subversion, could anyone get the complete package and zip it and host it somewhere please?

[QUOTE=ATH;138662]What parameters did you use to compile it?

I'm trying to use the Msys and the MinGW 64 bit to compile it on Windows x64, but I can't get it to work yet.[/QUOTE]

Apart from applying [url=http://www.math.jmu.edu/~martin/gmp-4.2.x-core2-port.tar.gz]this patch[/url] to GMP (Which really does speed things up significantly on Core 2 machines) I didn't use any special parameters. I just ran ./configure on a Core 2 Duo machine with 64-bit Debian Linux installed and let the script choose the parameters. I have had no success in trying to build GMP from Linux with the mingw cross-compilers.

Luigi: sorry I mis-read your question. No I didn't link with the GWNUM library.

Here are some binaries configured with --enable-asm-redc for "small" input numbers.

Windows (32bit), Core 2 Duo: [URL="http://www2.informatik.hu-berlin.de/~schoenbe/ecm621_win32_core2.zip"]ecm621_win32_core2.zip[/URL]
Linux (32bit), Pentium M: [URL="http://www2.informatik.hu-berlin.de/~schoenbe/ecm621_lin32_pentiumm.tar.gz"]ecm621_lin32_pentiumm.tar.gz[/URL]
Linux (64bit), Core 2 Duo: [URL="http://www2.informatik.hu-berlin.de/~schoenbe/ecm621_lin64_core2.tar.gz"]ecm621_lin64_core2.tar.gz[/URL]

Here is an Intel x86 Windows 64bit GMP-ECM 6.2 rc2 with GMP 4.2.2 compiled by [URL="http://www.mersenneforum.org/member.php?u=6952"]jbristow[/URL]. He emailed it to me and I said I could host it and post it in this thread:

[URL="http://www.hoegge.dk/mersenne/ecm62rc2_64bit.zip"]ecm62rc2_64bit.zip[/URL]

Exactly which Core2duo or Core2quad was it compiled on?

[quote=ATH;140573]Here is an Intel x86 Windows 64bit GMP-ECM 6.2 rc2 with GMP 4.2.2 compiled by [URL="http://www.mersenneforum.org/member.php?u=6952"]jbristow[/URL]. He emailed it to me and I said I could host it and post it in this thread:

[URL="http://www.hoegge.dk/mersenne/ecm62rc2_64bit.zip"]ecm62rc2_64bit.zip[/URL]

Exactly which Core2duo or Core2quad was it compiled on?[/quote]

It's a Q6600.

I'm not sure its a true 64bit compiled GMP-ECM. It crashed when it tried to allocate 2998Mb for a stage2, and I had 3.3Gb free. I'm using Windows XP 64-bit.

[quote=ATH;140660]I'm not sure its a true 64bit compiled GMP-ECM. It crashed when it tried to allocate 2998Mb for a stage2, and I had 3.3Gb free. I'm using Windows XP 64-bit.[/quote]

It is running happily right now on Windows Server 2003 R2, consuming 13GB in stage 2. This is a 16GB system, so there is still plenty of room for background processes. Perhaps that is your problem - not enough contiguous heap space?

- ben.

Pentium 4, Linux (32bit), --enable-asm-redc: [URL="http://www2.informatik.hu-berlin.de/~schoenbe/ecm621_lin32_pentium4.tar.gz"]ecm621_lin32_pentium4.tar.gz[/URL]

If I'm looking for an ecm binary for a 64-bit AMD Turion processor running Linux, would an ecm program optimized for an Athlon be a perfect fit for the Turion, a pretty good one, or a bad idea? I figured I'd ask before maybe downloading something that's sub-optimal. I'm not asking for people to find it for me, but four pages of text is a lot, so I'm basically asking for a casual opinion of whether or not I should scan the thread or make my own. Btw, a search for Turion in this thread just returns this exact post.

And if anybody actually HAS a Turion binary for 64-bit Linux, please post it here. :)

Athlon 64, Linux (64bit), --enable-asm-redc: [URL="http://www2.informatik.hu-berlin.de/~schoenbe/ecm621_lin64_athlon64.tar.gz"]ecm621_lin64_athlon64.tar.gz[/URL]

[quote=Yamato;142760]Athlon 64, Linux (64bit), --enable-asm-redc: [URL="http://www2.informatik.hu-berlin.de/~schoenbe/ecm621_lin64_athlon64.tar.gz"]ecm621_lin64_athlon64.tar.gz[/URL][/quote]
that would have been so helpful for me a month ago but unfortunately i got a Q6600 to replace it:smile:

None of the links seem to work. They all come up with an error. :sad:

Things seem to work for me. Check your browser settings (pop-up blocker, etc).

The error is a 404.

'www.hoegge.dk/mersenne' gives me a 403 error.

I've tried on another computer and the links still don't work.

[QUOTE=10metreh;149100]I've tried on another computer and the links still don't work.[/QUOTE]

Which file did you want? Files from post #75 and #84 still works, though the site could have been down when you tried it. If you are trying the links from earlier versions before 6.2rc1 I removed them.

My processor is a Pentium 4 1.7GHz. There is nothing for this processor.