The Second Megabit Drive - Page 19

Kosmaj · 2014-01-25, 23:57

Hi Carlos

I'll send you the file to 1700k next week. We haven't generated files past 1700k yet.

BTW, based on my exe times at 1570k, a 10k block on AVX, and a 4k block on non-AVX CPU are going to take about 2 months each. And the next FFT jamp is starting soon, at about 1600k, so exe times will go up.

pinhodecarlos · 2014-01-26, 01:09

So 5k chunks for AVX cores and 2.5k chunks for non-AVX although I prefer doubling but that's ok.
Don't forget that I will be stuck in the lab for the PhD for the next 4 years...lol
About the files I can download them from your website until 1700k, just didn't want to get busy joining the files but I will use them.
Is there a way to merge the files taking out the first line of each one?

Don't get busy generating further files, I will use the ones you have.

Thank you.

Carlos

PS( I have doubts if I should continue with k=5, 1st megabit drive or 2nd megabit drive, I will think about it over night sleep)

Kosmaj · 2014-01-26, 02:17

Assuming that you have a linux-like environment, open a terminal, go to the directory where you got all the files, and type:

Code:

> cat *.txt | grep -v ":" > all.txt

Regarding your doubts it's entirely up to you, but it will be nice if you first complete your outstanding reservations in the 1st drive and k=5. If you haven't started yet k=5 in the 4010-4090k range you can cancel it.

Quote:

Originally Posted by pinhodecarlos

Is there a way to merge the files taking out the first line of each one?
...
PS( I have doubts if I should continue with k=5, 1st megabit drive or 2nd megabit drive, I will think about it over night sleep)

pinhodecarlos · 2014-01-26, 08:03

k=5 in the 4010-4090k range is underway.

pinhodecarlos · 2014-01-26, 11:20

Quote:

Originally Posted by pinhodecarlos

k=5 in the 4010-4090k range is underway.

I posted the progress of that range.
Also, after a good night of sleep I decided to keep crunching k=5 so I will be moving the cores (AVX ones) running 1st and 2nd Megabit drives to there. The goal is to move the k=5 up to 6M while I do my PhD.
Also my friend in the lab will be donating one AVX core to me so next Monday I will reserve k=5 work for him. Probably I can get another one from another colleague student, I need to know which machine do they gave him.

Sorry for the inconvenience but sometimes we have doubts.

Carlos

pinhodecarlos · 2014-02-06, 22:12

1574-1576 complete, one prime on the last candidate of the file!!!

pinhodecarlos · 2014-02-06, 22:18

1530-1538 complete, no primes.

pinhodecarlos · 2014-02-06, 22:20

1540-1549 complete, 2 primes.

pinhodecarlos · 2014-02-10, 10:58

1549-1550 complete, no primes.

pinhodecarlos · 2014-02-17, 22:11

1538-1540 complete, no primes
This completes 1530-1550.

pinhodecarlos · 2014-02-20, 12:15

1576-1578 complete, no primes.

2014-01-26, 01:09	#200
pinhodecarlos "Carlos Pinho" Oct 2011 Milton Keynes, UK 3×17×97 Posts	So 5k chunks for AVX cores and 2.5k chunks for non-AVX although I prefer doubling but that's ok. Don't forget that I will be stuck in the lab for the PhD for the next 4 years...lol About the files I can download them from your website until 1700k, just didn't want to get busy joining the files but I will use them. Is there a way to merge the files taking out the first line of each one? Don't get busy generating further files, I will use the ones you have. Thank you. Carlos PS( I have doubts if I should continue with k=5, 1st megabit drive or 2nd megabit drive, I will think about it over night sleep) Last fiddled with by pinhodecarlos on 2014-01-26 at 01:35

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
The Third Megabit Drive	pinhodecarlos	Riesel Prime Search	135	2021-04-26 16:44
The First Megabit Drive	Kosmaj	Riesel Prime Search	373	2020-11-12 15:48
RPS 11th Drive: Search for more megabit primes	Kosmaj	Riesel Prime Search	186	2014-12-04 14:03
Any interest in yet another sub-megabit drive?	Kosmaj	Riesel Prime Search	20	2011-11-17 04:10
The 3rd RPS Drive: low-weight Ks for megabit prime	lsoule	Riesel Prime Search	140	2009-03-02 15:01

2014-01-25, 23:57	#199
Kosmaj Nov 2003 2·1,811 Posts	Hi Carlos I'll send you the file to 1700k next week. We haven't generated files past 1700k yet. BTW, based on my exe times at 1570k, a 10k block on AVX, and a 4k block on non-AVX CPU are going to take about 2 months each. And the next FFT jamp is starting soon, at about 1600k, so exe times will go up.

2014-01-26, 08:03	#202
pinhodecarlos "Carlos Pinho" Oct 2011 Milton Keynes, UK 3×17×97 Posts	k=5 in the 4010-4090k range is underway.