-   Twin Prime Search (
-   -   Archived sieve reservations and discussions (

Lennart 2010-05-12 00:23

[quote=gribozavr;214695]For P<65T I'm missing some Joshua2's factors:

Joshua2's 90T-120T file unpacked.

Lennart: can't connect to port 80, although that host replies to ping requests.

Sieving input: [URL][/URL] [URL][/URL]
480000-484999_11may2010.txt.gz [URL][/URL] [URL][/URL]
480000-484999_11may2010.txt.xz [URL][/URL] [URL][/URL] [URL][/URL] [URL][/URL]
485000-489999__11may2010.txt.gz [URL][/URL] [URL][/URL]
485000-489999__11may2010.txt.xz [URL][/URL] [URL][/URL]

LLR input (although I don't recommend doing any LLR now):

n_480000-480999__k_1_100000.txt [URL][/URL][/quote]

I'll start 10T-15T


Mini-Geek 2010-05-12 01:03

[quote=Historian;214742]I see. What's your p/sec sieve rate,[/quote]
Two separate trials, each trying to use 3 threads:[code]Sieve complete: 120000000000000 <= p < 120000010000000
Found 0 factors
Elapsed time: 43.92 sec. (42.86 init + 1.06 sieve) at 9868951 p/sec.
Processor time: 33.48 sec. (31.20 init + 2.28 sieve) at 4596498 p/sec.
Average processor utilization: 0.73 (init), 2.15 (sieve)

Sieve complete: 120000000000000 <= p < 120000100000000
Found 2 factors
Elapsed time: 102.95 sec. (95.19 init + 7.77 sieve) at 12895164 p/sec.
Processor time: 53.17 sec. (30.77 init + 22.41 sieve) at 4469244 p/sec.
Average processor utilization: 0.32 (init), 2.89 (sieve)[/code]So roughly 4.5M p/sec per core (32-bit OS).
[quote=Historian;214742]and what processor are you using?[/quote]
Quad-core Intel i5-750 running at 2807 MHz on 32-bit Windows XP. 64-bit should be drastically faster, perhaps twice as fast, but I don't know exactly how much faster.

If I had a 64-bit OS installed and more RAM available (2GB to share between Windows, Firefox, PRPnet server and clients, and any other apps I want to run doesn't really leave 1GB totally free for a sieving effort; and the OS bit difference means I'd be suboptimal), I'd participate more in this sieve.

Oddball 2010-05-12 08:38

I don't have enough memory to participate in the sieve, but I'm still interested in knowing the sieve rate[/QUOTE]
You might want to split up the sieve file into three parts (485000-486999, 487000-488399, 488400-490000), like I did. I'm getting ~13M/sec on a 3.2 GHz Pentium 4 for one part. Divide that by 3, and you get 4-5M/sec for the whole range.

Historian 2010-05-12 19:00

[QUOTE=Oddball;214785]You might want to split up the sieve file into three parts (485000-486999, 487000-488399, 488400-490000), like I did.[/QUOTE]
Good idea. I'll take a quickie:

Reserving 2500-2525, for the 485K-490K part.

Historian 2010-05-12 21:27

2500-2525G complete:


18,000+ factors found

Lennart 2010-05-12 22:18

10T-15T are done !



gribozavr 2010-05-13 00:06

Historian, Lennart: downloaded and verified your files.

agent1 2010-05-13 19:39

reserving 2525-2600 range2

Oddball 2010-05-14 01:49

[QUOTE=agent1;214966]reserving 2525-2600 range2[/QUOTE]
Welcome to TPS! I've added your range to the first post, and I'll finish up my 500G range sometime by tomorrow.

mdettweiler 2010-05-14 02:30

Reserving 2600G-2700G for n=485K-490K. :smile:

mdettweiler 2010-05-14 16:33

[quote=mdettweiler;214991]Reserving 2600G-2700G for n=485K-490K. :smile:[/quote]
Done, 68755 factors: [URL][/URL]

Reserving 25T-26T for n=480K-485K.

All times are UTC. The time now is 22:06.

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2022, Jelsoft Enterprises Ltd.