mersenneforum.org

mersenneforum.org (https://www.mersenneforum.org/index.php)
-   Twin Prime Search (https://www.mersenneforum.org/forumdisplay.php?f=65)
-   -   Archived sieve reservations and discussions (https://www.mersenneforum.org/showthread.php?t=13389)

Lennart 2010-05-11 02:01

[URL]http://sites.google.com/site/kenscode/prime-programs[/URL]

Here you have the site. Here you will find the latest tpsieve files.


Lennart

Lennart 2010-05-11 02:56

[URL]http://uwin.mine.nu/TPS/480000-485000/factorupload/tpsfactors_65T-T66T.txt.zip[/URL]


65T-66T done

Lennart

Oddball 2010-05-11 03:40

The first two posts are fully updated now - range reservations up to p=66T, individual stats (joshua2 is on top), and the old link to download TPSieve has been replaced with the new one.

Gribozavr, could you:

- confirm that you have the factor files for all p<66T
- upload the sieve file containing only small k's (k<100,000) for 480000<n<481000
- upload the entire k<10M file for 480000<n<485000 and for 485000<n<500000

gribozavr 2010-05-11 16:36

For P<65T I'm missing some Joshua2's factors:
10T-15T
25T-35T
45T-65T

Joshua2's 90T-120T file unpacked.

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

Sieving input:

480000-484999_11may2010.zip [url]http://www.sendspace.com/file/srou2c[/url] [url]http://www.rapidspread.com/file.jsp?id=wvkxjrbevp[/url]
480000-484999_11may2010.txt.gz [url]http://www.sendspace.com/file/on86ar[/url] [url]http://www.rapidspread.com/file.jsp?id=pzd7a1yjmu[/url]
480000-484999_11may2010.txt.xz [url]http://www.sendspace.com/file/95gw4b[/url] [url]http://www.rapidspread.com/file.jsp?id=kb5kdg4oen[/url]

485000-489999__11may2010.zip [url]http://www.sendspace.com/file/sm8la6[/url] [url]http://www.rapidspread.com/file.jsp?id=zdxx1pen6k[/url]
485000-489999__11may2010.txt.gz [url]http://www.sendspace.com/file/bwnhkv[/url] [url]http://www.rapidspread.com/file.jsp?id=v5tvvxal7a[/url]
485000-489999__11may2010.txt.xz [url]http://www.sendspace.com/file/550otw[/url] [url]http://www.rapidspread.com/file.jsp?id=x9adyjrap9[/url]

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

n_480000-480999__k_1_100000.txt [url]http://www.sendspace.com/file/pxytxr[/url]

gribozavr 2010-05-11 19:47

Lennart: downloaded and verified your file.

Lennart 2010-05-11 19:51

[quote=gribozavr;214716]Lennart: downloaded and verified your file.[/quote]
:tu:

Lennart

Historian 2010-05-11 23:14

[QUOTE=gribozavr;214695]
Joshua2's 90T-120T file unpacked.
[/QUOTE]
Does this mean that all factors between 90T and 120T have been removed?

Also, does anyone have some benchmarks on the sieve and how much memory is needed to run it?

Mini-Geek 2010-05-11 23:25

[quote=Historian;214734]Also, does anyone have some benchmarks on the sieve and how much memory is needed to run it?[/quote]
With the last sieve file I've run, (I'd assume there are no huge changes in this area due to the latest factor removals, though I don't know my last exact depth) it takes just under 1GB of memory to run it. Note that tpsieve is multithreaded, so you could run it on all the cores you want while using only 1GB of memory (instead of requiring 1GB of memory per core).

gribozavr 2010-05-11 23:48

[QUOTE=Historian;214734]Does this mean that all factors between 90T and 120T have been removed?[/QUOTE]

Yes, 90T-120T should be added to the first post as completed.

Historian 2010-05-11 23:52

[QUOTE=Mini-Geek;214736]With the last sieve file I've run, (I'd assume there are no huge changes in this area due to the latest factor removals, though I don't know my last exact depth) it takes just under 1GB of memory to run it. Note that tpsieve is multithreaded, so you could run it on all the cores you want while using only 1GB of memory (instead of requiring 1GB of memory per core).[/QUOTE]
I see. What's your p/sec sieve rate, and what processor are you using?

I don't have enough memory to participate in the sieve, but I'm still interested in knowing the sieve rate so I can estimate how much computing time was spent on this project.

gribozavr 2010-05-12 00:06

Much "32-bit CPU time" was spent on sieving. Now, with tpsieve and a multicore 64-bit CPU required time has dropped in... I don't know for sure, but my rough estimate is 3..7 times.


All times are UTC. The time now is 18:14.

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