34-34.2 upto 2^62

Khemikal796 · 2003-06-19, 20:00

I have got around to doing some benchies on my Duron. If everything runs fine throughout the summer I should have no problem finishing off this range upto 62 bits. There is even a slight possibility I will end up with a new 2.4C system running on these to finish them off a little faster before I turn it loose on LL testing. Just gotta wait and see if I can get mobo and ram for this lonely chip.

Khemikal796 · 2003-07-25, 12:42

I got 1/20th of this range done so far with the 800 Duron. I have recently switched over to the 2.4C I was talking about. I finished the 2 last exponents I had reserved on primenet and when I get home today will move it over to doing 34.02-34.2

garo · 2003-07-25, 17:19

Not a good idea. P4s suck at factoring to bits below 62. Your P4 2.4 will be less than twice as fast as the Duron. Better stick with the Duron.

Khemikal796 · 2003-07-25, 21:04

well wish i could, but i dont have the psu, hdds, or a case to put it in. If i were to actually keep this duron up and running i would need a new mobo too because it was crapping out on me. Thats why i upgraded to the P4. Ill run some times and see if this is true. I can always try factoring any ranges that have been done to the P4 range then. I wanna do some factoring if you havent noticed yet :)

Khemikal796 · 2003-07-25, 21:26

Did the trials and it is faster than the old duron. With pc133 ram and the duron at 800 MHz i was getting times of about 9.xx secs for 500 iterations. With the new 2.4C and ddr at 400MHz im getting times of 6.xx seconds. So i will continue doing this range till it is done.

garo · 2003-07-26, 00:03

Well, it's your hardware and it is your choice. But if I were you I would think about doing LL testing as the P4 is 4 times faster than the Duron as opposed to being 1.5 times as fast in factoring.

If you really want to factor, I would factor these numbers to 2^68 as the P4 really shines at the higher bit levels. You can leave the ones you have already done to 2^62 and just change your Factoroverride to 68 or if that is too much for you even 66 but make it larger than 62. Please :( :( :(

And FSB and memory speed does not matter in factoring but a lot in LL testing.

lycorn · 2003-07-27, 06:44

Quote:

Originally Posted by garo

Not a good idea. P4s suck at factoring to bits below 62.

AFAIR, SSE2 optimizations only kick in at 64 bits. So, P4s shouldn´t be, from an overall througput maximization perspective, used for TFing below that bit level.
Just a hint, I don´t mean to be pushy... :)

Khemikal796 · 2003-08-26, 03:56

I have finished up to 34.05 and will be sending this in to George. I am going to be going back to LL. Please next time just let someone do what they want to do and not tell them like you told me they were doing something wrong. I mentioned I knew I could be doing better at LL but I still wanted to help LMH. I felt very unappreciated for just trying to give you cycles. From what I understood you wanted everything up to 62 bit so I took the box I had available and used it to help. It might not have been the best use of my machine, but its better than letting the cycles go to waste.

Im not sure how much I enjoyed my time here, but I know I didnt totally hate it. I might come back some day when I have some machines that will run a little 'better' at the goal of the project.

Thanks for the time I had here.

garo · 2003-08-26, 18:47

I am sorry if you felt we were being pushy. We were just trying to help you optimize the use of your machine. Most DC participants like to use their machine for the job that is optimum for them and we felt you wanted the same too. We only wanted to help you. I am sorry you did not feel that way.

2003-06-19, 20:00	#1
Khemikal796 Feb 2003 100101₂ Posts	34-34.2 upto 2^62 I have got around to doing some benchies on my Duron. If everything runs fine throughout the summer I should have no problem finishing off this range upto 62 bits. There is even a slight possibility I will end up with a new 2.4C system running on these to finish them off a little faster before I turn it loose on LL testing. Just gotta wait and see if I can get mobo and ram for this lonely chip.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
46 - 46.5 upto 2^61	antiroach	Lone Mersenne Hunters	1	2005-06-17 06:22
51.0-51.5M upto 2^61	antiroach	Lone Mersenne Hunters	1	2005-05-13 05:34
59.0 - 59.8 upto 2^61	antiroach	Lone Mersenne Hunters	3	2005-03-28 14:14
66-69 upto 2^59	antiroach	Lone Mersenne Hunters	1	2004-01-18 23:32
77-78M upto 2^61	antiroach	Lone Mersenne Hunters	5	2003-11-30 23:53

2003-07-25, 12:42	#2
Khemikal796 Feb 2003 37 Posts	I got 1/20th of this range done so far with the 800 Duron. I have recently switched over to the 2.4C I was talking about. I finished the 2 last exponents I had reserved on primenet and when I get home today will move it over to doing 34.02-34.2

2003-07-25, 17:19	#3
garo Aug 2002 Termonfeckin, IE 2⁴×173 Posts	Not a good idea. P4s suck at factoring to bits below 62. Your P4 2.4 will be less than twice as fast as the Duron. Better stick with the Duron.

2003-07-25, 21:04	#4
Khemikal796 Feb 2003 37₁₀ Posts	well wish i could, but i dont have the psu, hdds, or a case to put it in. If i were to actually keep this duron up and running i would need a new mobo too because it was crapping out on me. Thats why i upgraded to the P4. Ill run some times and see if this is true. I can always try factoring any ranges that have been done to the P4 range then. I wanna do some factoring if you havent noticed yet :)

2003-07-25, 21:26	#5
Khemikal796 Feb 2003 37 Posts	Did the trials and it is faster than the old duron. With pc133 ram and the duron at 800 MHz i was getting times of about 9.xx secs for 500 iterations. With the new 2.4C and ddr at 400MHz im getting times of 6.xx seconds. So i will continue doing this range till it is done.

2003-07-26, 00:03	#6
garo Aug 2002 Termonfeckin, IE 2⁴·173 Posts	Well, it's your hardware and it is your choice. But if I were you I would think about doing LL testing as the P4 is 4 times faster than the Duron as opposed to being 1.5 times as fast in factoring. If you really want to factor, I would factor these numbers to 2^68 as the P4 really shines at the higher bit levels. You can leave the ones you have already done to 2^62 and just change your Factoroverride to 68 or if that is too much for you even 66 but make it larger than 62. Please :( :( :( And FSB and memory speed does not matter in factoring but a lot in LL testing.

2003-08-26, 03:56	#8
Khemikal796 Feb 2003 45₈ Posts	I have finished up to 34.05 and will be sending this in to George. I am going to be going back to LL. Please next time just let someone do what they want to do and not tell them like you told me they were doing something wrong. I mentioned I knew I could be doing better at LL but I still wanted to help LMH. I felt very unappreciated for just trying to give you cycles. From what I understood you wanted everything up to 62 bit so I took the box I had available and used it to help. It might not have been the best use of my machine, but its better than letting the cycles go to waste. Im not sure how much I enjoyed my time here, but I know I didnt totally hate it. I might come back some day when I have some machines that will run a little 'better' at the goal of the project. Thanks for the time I had here.

2003-08-26, 18:47	#9
garo Aug 2002 Termonfeckin, IE AD0₁₆ Posts	I am sorry if you felt we were being pushy. We were just trying to help you optimize the use of your machine. Most DC participants like to use their machine for the job that is optimum for them and we felt you wanted the same too. We only wanted to help you. I am sorry you did not feel that way.