LLR 3.8.4 development version is available!

Jean Penné · 2010-11-02, 20:25

Hi,
Since the 25/10/10, the Windows and Linux binaries of the development version 3.8.4 of LLR can be dowloaded on my development directory :

http://jpenne.free.fr/Development/

And thanks to Iain Bethune, I uploaded today the Mac Intel binary on the same directory.

This version is identical to 3.8.2, apart from beeing linked with gwnum V26.4

I will upload the corresponding sources when "official" gwnum V26.4 will be released by George.

Regards,
Jean

MooMoo2 · 2010-11-02, 22:25

Is it just me, or is this version a bit slower than version 3.8.2? For the 80K FFT length, I'm getting 1.055 sec/iteration, but I'm getting 1.048 sec/iteration using version 3.8.2.

If there are no problems with version 3.8.2 (other than the bugs on this list: http://www.mersenneforum.org/showpos...81&postcount=2 ), I'll continue to use it instead of switching to 3.8.4. I'm only testing base 2 numbers on an AMD computer using Windows 7, so those bugs shouldn't affect me.

em99010pepe · 2010-11-03, 19:58

Quote:

Originally Posted by MooMoo2

Is it just me, or is this version a bit slower than version 3.8.2? For the 80K FFT length, I'm getting 1.055 sec/iteration, but I'm getting 1.048 sec/iteration using version 3.8.2.

If there are no problems with version 3.8.2 (other than the bugs on this list: http://www.mersenneforum.org/showpos...81&postcount=2 ), I'll continue to use it instead of switching to 3.8.4. I'm only testing base 2 numbers on an AMD computer using Windows 7, so those bugs shouldn't affect me.

I don't see under Core 2 machines a speed up when using version 3.8.2 when testing base 2 numbers. After a few hours of testing I think both versions have the same speed but I could be wrong.

Ken_g6 · 2010-11-14, 16:57

Henry asked this in the v3.8.2 thread:

Quote:

Originally Posted by henryzz

Is there any plans for a 64-bit LLR? According to

pfgw is expected to get some improvement from this with v26.2. Another speedup if available would be great.

I didn't see any reply, and I'm wondering the same thing.

em99010pepe · 2010-11-14, 17:32

I tested some riesel numbers on base 2 with pfgw 64 bit and timings were worst than LLR 3.8.4 32 bits.

2010-11-02, 20:25	#1
Jean Penné May 2004 FRANCE 571 Posts	LLR 3.8.4 development version is available! Hi, Since the 25/10/10, the Windows and Linux binaries of the development version 3.8.4 of LLR can be dowloaded on my development directory : http://jpenne.free.fr/Development/ And thanks to Iain Bethune, I uploaded today the Mac Intel binary on the same directory. This version is identical to 3.8.2, apart from beeing linked with gwnum V26.4 I will upload the corresponding sources when "official" gwnum V26.4 will be released by George. Regards, Jean

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
LLR Development Version 3.8.9 released.	Jean Penné	Software	6	2012-05-09 23:44
LLR development version 3.8.7 is available!	Jean Penné	Software	39	2012-04-27 12:33
LLR 3.8.6 Development version	Jean Penné	Software	0	2011-06-16 20:05
LLR 3.8.5 Development version	Jean Penné	Software	6	2011-04-28 06:21
PowerPC development platform	nuggetprime	Hardware	2	2011-01-06 18:37

2010-11-02, 22:25	#2
MooMoo2 Aug 2010 3·5·41 Posts	Is it just me, or is this version a bit slower than version 3.8.2? For the 80K FFT length, I'm getting 1.055 sec/iteration, but I'm getting 1.048 sec/iteration using version 3.8.2. If there are no problems with version 3.8.2 (other than the bugs on this list: http://www.mersenneforum.org/showpos...81&postcount=2 ), I'll continue to use it instead of switching to 3.8.4. I'm only testing base 2 numbers on an AMD computer using Windows 7, so those bugs shouldn't affect me.

2010-11-14, 17:32	#5
em99010pepe Sep 2004 2×5×283 Posts	I tested some riesel numbers on base 2 with pfgw 64 bit and timings were worst than LLR 3.8.4 32 bits.