Data needed... Please Help!

guido72 · 2003-03-29, 16:21

Hi all!
This is Guido, from Italy.
Me and Luigi are keeping up-to-date the italian G.I.M.P.S. mirror and working on our "Team_Italia" website (It's time to shake with fear, TPR-ers :-) )
Well, the more we go on, the bigger the eponents, the more the patience needed... Above all here, where the hardware is not so updated. Many People may be interested in factoring job, which doesn't require state-of-the-art machines. I've prepared a short script similar to one you may find at http://www.mersenne.org/bench.htm. So, now
I'd just need some data about Time per Iteration for nothing else but factoring jobs, of all kind of CPUs.
To get reliable data, I suggest to set at least at "1000" the iterations before screen outputs option. Then try to see if this Time is always the same the deeper you go with factorization. You may send your benchies to me. My e-mail address is loguidoNOSPAM@katamail.com
Thank you all in advance!
Regards
Guido

guido72 · 2003-04-01, 14:49

Hi!
Nobody may help me? :(

:(

lycorn · 2003-04-01, 17:54

Hi Guido!

The following values were obtained with an Athlon XP2200+, using PC2100 DDR, running at stock speed. Job was TF in the 77-77,5M range, from 59 to 60 bits.
For 2000 iterations between screen outputs, the time taken would vary between 13.9 and 14.9 secs (if there were values outside this interval, they were very rare). These values have been similar for factoring between 58 and 59 bits
On a PIII-700, 1000 iterations in the 77.5-78M range, but from 58 to 59 bits, usually take between 25.1 and 26.8 secs. I had already done TFing using that machine in the 46-47M and 78-79M ranges, and did not spot any significant difference.

Diogo (lycorn)

guido72 · 2003-04-01, 18:24

Hi!
As first answerer you won.... Nothing! ;)
Jokes apart: Thanks a lot!
Come on boys! I need data!
Regards from Italy.
Guido

garo · 2003-04-01, 18:51

[code:1]
Time taken to complete 14366959 to 0.98%

To Bit On PII 450 On P4 2533 Improv On 1333TB Improv

59 6 2.5 2.4 1.85 3.25
60 12 5 2.4 3.7 3.25
61 24 10 2.4 7.4 3.25
62 48 20 2.4 14.8 3.25
63 170 37.5 4.5 51 3.33
64 340 75 4.5 102 3.33
65 1540 180 8.5 480 3.21
[/code:1]

guido72 · 2003-04-01, 20:33

Thank you Garo!

guido72 · 2003-04-01, 20:35

If I'll able to collect a good database, we should use these data to improve the Benchies pages of Gimps website as well... So let's post your benchies!
See you tomorrow and Good Night!
Guido

markr · 2003-04-02, 01:39

Hi Guido! I hope this helps, although I'm not sure if it's exactly in the form you asked for. I wanted to find out how long various exponents would take, so this is how I collected it.

[code:1]exponent bits from* bits to time (m) machine

3400xxxx 58 60 6:36 Athlon 900

3400xxxx 58 60 54:34 P166

3400xxxx 58 60 7:57 P1800

47000xxx 58 60 5:38 P1800

42057373 65 66 325 P1800
66 67 649
67 68 1300
P-1 stage 1 2627

* bits from: exponent already factored to this many bits.[/code:1]

The first four rows have average times over several exponents. The other times are estimated from the time taken for most of the bit-level.

guido72 · 2003-04-02, 06:48

Wow! That's the power of the net!
Thank you Mark! Actually I've a Formula George gave me some time ago to calculate the number of iterations requested for each stage of factorization, so I should be able to obtain the Time per iteration even from your data. But, if you all may give me this information as well, I mean the time per iteration during a factorization job, we may check George's formula as well. Anyway, Thank you very much once again, Mark and regards from Italy.
Guido

Reboot It · 2003-04-02, 13:01

Quote:

Originally Posted by guido72

I've a Formula George gave me some time ago to calculate the number of iterations requested for each stage of factorization...

Could you post that formula, please, or perhaps a link to it? It would be interesting.

garo · 2003-04-02, 18:36

markr,
You really should not be doing ANY factoring under 62 bits on a P4. It's wway suboptimal. Do it to 62 or even 64 bits on the Athlon and the rest on the P4.

2003-03-29, 16:21	#1
guido72 Aug 2002 Rovereto (Italy) 9F₁₆ Posts	Data needed... Please Help! Hi all! This is Guido, from Italy. Me and Luigi are keeping up-to-date the italian G.I.M.P.S. mirror and working on our "Team_Italia" website (It's time to shake with fear, TPR-ers :-) ) Well, the more we go on, the bigger the eponents, the more the patience needed... Above all here, where the hardware is not so updated. Many People may be interested in factoring job, which doesn't require state-of-the-art machines. I've prepared a short script similar to one you may find at http://www.mersenne.org/bench.htm. So, now I'd just need some data about Time per Iteration for nothing else but factoring jobs, of all kind of CPUs. To get reliable data, I suggest to set at least at "1000" the iterations before screen outputs option. Then try to see if this Time is always the same the deeper you go with factorization. You may send your benchies to me. My e-mail address is loguidoNOSPAM@katamail.com Thank you all in advance! Regards Guido

2003-04-01, 18:24	#4
guido72 Aug 2002 Rovereto (Italy) 3·53 Posts	Thanks a lot! Hi! As first answerer you won.... Nothing! ;) Jokes apart: Thanks a lot! Come on boys! I need data! Regards from Italy. Guido

2003-04-02, 06:48	#9
guido72 Aug 2002 Rovereto (Italy) 3×53 Posts	From Australia to Italy... 0 seconds... Wow! That's the power of the net! Thank you Mark! Actually I've a Formula George gave me some time ago to calculate the number of iterations requested for each stage of factorization, so I should be able to obtain the Time per iteration even from your data. But, if you all may give me this information as well, I mean the time per iteration during a factorization job, we may check George's formula as well. Anyway, Thank you very much once again, Mark and regards from Italy. Guido

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Corrupted data?	Oops!	Information & Answers	2	2013-10-22 03:48
GPU TF vs DC/LL data	bcp19	GPU to 72	0	2011-12-02 16:41
Data available?	Prime95	LMH > 100M	10	2007-06-22 23:55
P3 data needed	Prime95	Software	13	2003-10-02 04:10
New "Data" forum might be needed	GP2	Data	8	2003-09-13 02:34

2003-04-01, 14:49	#2
guido72 Aug 2002 Rovereto (Italy) 9F₁₆ Posts	Hi! Nobody may help me? :( :(

2003-04-01, 17:54	#3
lycorn Sep 2002 Oeiras, Portugal 2⁶·23 Posts	Hi Guido! The following values were obtained with an Athlon XP2200+, using PC2100 DDR, running at stock speed. Job was TF in the 77-77,5M range, from 59 to 60 bits. For 2000 iterations between screen outputs, the time taken would vary between 13.9 and 14.9 secs (if there were values outside this interval, they were very rare). These values have been similar for factoring between 58 and 59 bits On a PIII-700, 1000 iterations in the 77.5-78M range, but from 58 to 59 bits, usually take between 25.1 and 26.8 secs. I had already done TFing using that machine in the 46-47M and 78-79M ranges, and did not spot any significant difference. Diogo (lycorn)

2003-04-01, 18:51	#5
garo Aug 2002 Termonfeckin, IE 2²·691 Posts	[code:1] Time taken to complete 14366959 to 0.98% To Bit On PII 450 On P4 2533 Improv On 1333TB Improv 59 6 2.5 2.4 1.85 3.25 60 12 5 2.4 3.7 3.25 61 24 10 2.4 7.4 3.25 62 48 20 2.4 14.8 3.25 63 170 37.5 4.5 51 3.33 64 340 75 4.5 102 3.33 65 1540 180 8.5 480 3.21 [/code:1]

2003-04-01, 20:33	#6
guido72 Aug 2002 Rovereto (Italy) 3·53 Posts	Thank you Garo!

2003-04-01, 20:35	#7
guido72 Aug 2002 Rovereto (Italy) 237₈ Posts	If I'll able to collect a good database, we should use these data to improve the Benchies pages of Gimps website as well... So let's post your benchies! See you tomorrow and Good Night! Guido

2003-04-02, 01:39	#8
markr "Mark" Feb 2003 Sydney 1075₈ Posts	Hi Guido! I hope this helps, although I'm not sure if it's exactly in the form you asked for. I wanted to find out how long various exponents would take, so this is how I collected it. [code:1]exponent bits from* bits to time (m) machine 3400xxxx 58 60 6:36 Athlon 900 3400xxxx 58 60 54:34 P166 3400xxxx 58 60 7:57 P1800 47000xxx 58 60 5:38 P1800 42057373 65 66 325 P1800 66 67 649 67 68 1300 P-1 stage 1 2627 * bits from: exponent already factored to this many bits.[/code:1] The first four rows have average times over several exponents. The other times are estimated from the time taken for most of the bit-level.

2003-04-02, 18:36	#11
garo Aug 2002 Termonfeckin, IE 5314₈ Posts	markr, You really should not be doing ANY factoring under 62 bits on a P4. It's wway suboptimal. Do it to 62 or even 64 bits on the Athlon and the rest on the P4.