Team drive #9 k=1005-2000 n=50K-350K - Page 12

gd_barnes · 2009-01-09, 10:21

Quote:

Originally Posted by kar_bon

try

Code:

type *.txt >all.txt

this will 'copy' (actually list) all *.txt-files in one file called 'all.txt' without the $1A-value at the end!

Thanks Karsten. I'll try that.

Gary

Flatlander · 2009-01-09, 13:47

Quote:

Originally Posted by gd_barnes

Added a quad to the fun. I will likely add 3 more cores later today.

...

Gary

In that case, you will reach a score 0f 1000,000 real quick.

http://stats4.free-dc.org/stats.php?...name=gd_barnes

IronBits · 2009-01-09, 13:58

copy /b part1 + part2 + part3 file.ext

Brrrrrrrrrrr Carlos!

henryzz · 2009-01-09, 14:00

this drive is convicing me that we need a kind of score system for llrnet results
with this drive i have jumped from 2 primes to over 10 with very little effort
also my number of pairs tested has rocketed

mdettweiler · 2009-01-09, 21:10

Quote:

Originally Posted by gd_barnes

Max, it might be best to go ahead and process the results in ranges of 100k (actually ~50 k's), i.e. k=1005-1100, 1100-1200, etc. As soon as you see that the first k/n pair is k>=1101, then you can go ahead with it.

As soon as I can after getting the file from you, I'll start compiling a list of primes and finders. THAT should be interesting! lol

Okay. I see that the first k/n pair remaining in the server is for k=1107, so it's all set for me to process the results any time now.

One quick question: would you prefer the results for each 100-k range sorted by k or by n within that k-range? My scripts are designed to sort by n, but I should be able to modify them to sort by k if need be. Of course, even if I do end up sorting them by n, I'd still have to re-sort the sieve file...so, I'm going to end up having to change *something* either way. Though, of course, either modification should be pretty easy, so I'm fine with either way.

gd_barnes · 2009-01-09, 23:19

Quote:

Originally Posted by mdettweiler

Okay. I see that the first k/n pair remaining in the server is for k=1107, so it's all set for me to process the results any time now.

One quick question: would you prefer the results for each 100-k range sorted by k or by n within that k-range? My scripts are designed to sort by n, but I should be able to modify them to sort by k if need be. Of course, even if I do end up sorting them by n, I'd still have to re-sort the sieve file...so, I'm going to end up having to change *something* either way. Though, of course, either modification should be pretty easy, so I'm fine with either way.

Sort by k primary and n secondary. I'll be matching them with the original results, which are sorted by k then n. Also, the primes listing will need to be by k then n.

Gary

gd_barnes · 2009-01-09, 23:27

Guys, don't get too excited about this score thing for all the results and primes for this effort.

The scoring should be far lower for each pair and prime. Actually, it needs to be the same as calculated on the top-5000 site. That is:

A result at n=600K should score 36 times as much as a result at n=100K.

A prime at n=600K should score 216 times as much as a prime at n=100K.

That is correct, it takes 216 times as much CPU time to find a prime at n=600K vs. 100K and 36 times as much CPU time to process a result at n=600K vs. 100K.

That said, don't pull any machines off. I just wanted to point out that while this is a fun effort, it shouldn't score any higher than other efforts for the amount of CPU time expended.

Gary

glennpat · 2009-01-10, 02:31

I added one core for a couple of days for a little of this fast fun.

mdettweiler · 2009-01-10, 05:36

Quote:

Originally Posted by gd_barnes

Sort by k primary and n secondary. I'll be matching them with the original results, which are sorted by k then n. Also, the primes listing will need to be by k then n.

Gary

Okay, thanks--I'll get the results processed tomorrow (probably late afternoon or evening). Feel free to bug me if they're still not done by Sunday morning.

gd_barnes · 2009-01-10, 06:23

Quote:

Originally Posted by mdettweiler

Okay, thanks--I'll get the results processed tomorrow (probably late afternoon or evening). Feel free to bug me if they're still not done by Sunday morning.

By mid-day Sunday, we'll be done up to k=1200. If you wait until then, you may as well do k=1000-1200. lol

gd_barnes · 2009-01-10, 06:32

Moved discussion about the weather to a separate thread.

2009-01-09, 13:58	#124
IronBits I ♥ BOINC! Oct 2002 Glendale, AZ. (USA) 10001011001₂ Posts	copy /b part1 + part2 + part3 file.ext Brrrrrrrrrrr Carlos! Last fiddled with by IronBits on 2009-01-09 at 13:58

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Team drive #10 k=1400-2000 n=500K-1M	gd_barnes	No Prime Left Behind	61	2013-01-30 16:08
Team drive #12 k=2000-3000 n=50K-425K	gd_barnes	No Prime Left Behind	96	2012-02-19 03:53
Team drive #11 k=2000-3000 n=425K-600K	gd_barnes	No Prime Left Behind	42	2010-11-19 10:42
Team drive #8 k=1400-2000 n=350K-500K	gd_barnes	No Prime Left Behind	101	2009-04-08 02:11
Sieving drive for k=1005-2000 n=200K-500K	gd_barnes	No Prime Left Behind	118	2009-01-17 16:05

2009-01-09, 14:00	#125
henryzz Just call me Henry "David" Sep 2007 Cambridge (GMT/BST) 2³·3·5·7² Posts	this drive is convicing me that we need a kind of score system for llrnet results with this drive i have jumped from 2 primes to over 10 with very little effort also my number of pairs tested has rocketed

2009-01-09, 23:27	#128
gd_barnes May 2007 Kansas; USA 3³·5·7·11 Posts	Guys, don't get too excited about this score thing for all the results and primes for this effort. The scoring should be far lower for each pair and prime. Actually, it needs to be the same as calculated on the top-5000 site. That is: A result at n=600K should score 36 times as much as a result at n=100K. A prime at n=600K should score 216 times as much as a prime at n=100K. That is correct, it takes 216 times as much CPU time to find a prime at n=600K vs. 100K and 36 times as much CPU time to process a result at n=600K vs. 100K. That said, don't pull any machines off. I just wanted to point out that while this is a fun effort, it shouldn't score any higher than other efforts for the amount of CPU time expended. Gary

2009-01-10, 02:31	#129
glennpat May 2007 Minnesota USA 7² Posts	I added one core for a couple of days for a little of this fast fun.

2009-01-10, 06:32	#132
gd_barnes May 2007 Kansas; USA 3³×5×7×11 Posts	Moved discussion about the weather to a separate thread.