Sieving discussion thread

[KEP]

Hi

I was contacted by jmblazek asking for my support doing sieve, so now I'm asking can the range 1120 T - 1170 T (approximately 10 days of work), be reserved for me? Also which file do I have to sieve from, the one in this thread, or the one in the tutorial site?

Regards!

[Skligmund]

It actually doesn't matter which. I think the one in the tutorial might be a little smaller, so I'd go that route (corrrect me if I'm wrong).

If my upload was faster, I would let you download mine, as it is a few megs smaller yet.

[KEP]

Originally Posted by Skligmund

It actually doesn't matter which. I think the one in the tutorial might be a little smaller, so I'd go that route (corrrect me if I'm wrong).

If my upload was faster, I would let you download mine, as it is a few megs smaller yet.

Thanks, I've actually also decided to use that file from the tutorial, it was updated according to the time table in Windows, on January 19 2007, and it was approximately 10 megs smaller, so I'll stick with that one... thanks for your replys, happy sieving everyone :)

[pacionet]

I tried to start sieve n=500,000 for a little to test if all works. (0-300G)
NewPgen has successfully generated a PROTHGEN.TMP file of 5.58 GB, but after that , a windows popup appear with an error caused by the application (when NewPGen, after generating PROTHGEN.TMP, go to the next step).

Is it normal ?

[thommy]

no, that is not normal.
if you are on fat32 no files over 4GB or so are supportet. try it again with a ntfs structure. or split the ranges.

[pacionet]

No I am on NTFS, and the file PROTHGEN.TMP is still there (5.58 GB).

Probably the range is too large and this cause this error ? Is it better to split it ?

[pacionet]

Ok I have received this answer from Paul Jobling (NewPGen author):

Hi Andrea,
[...]

> Now we are switching to exponent n=333,333 and after,
> probably, n=500,000.
>
> I began to sieve n=500,000 from 0 to 300G. NewPGen has created a file,
> PROTHGEN.TMP of 5.58GB sieving all k to 1 billion, but immediately
> before start next step, it crashes (windows popup error saying there
> is an error caused by newpgen). The file PROTHGEN.TMP is still there.
>

[...]

> Is the range too large ? I have to split it ?

The space required to store those numbers in a bitmap is too large: 34 GB
are required. To store them in a list is also too large: if the file is 5.5
GB, I would imagine that it contains about 330 million different k values,
so the space required is 1.2 GB.

You will have to use a smaller range of k, unless you can find a PC with
more RAM, and simply load the file PROTHGEN.TMP into that.

Regards, and good luck,

Paul.

So MooMooo I'll not able to sieve the entire range, I am sorry.
I'll make some tests in these days and I'll let you know the maximum range of K I can reserve.

[jmblazek]

Why not do 0-207G? From biwema in a previous post in a different thread:

500000
FFT size: 48K
Time 1 test: 500 s
Optimal sieve bits: 57.5
Tests per twin: 23.7M
Twin every G: 90
90% chance after G: 207
CPU years per twin: 376

Actually, biwema suggests 250G...
http://www.mersenneforum.org/showpos...4&postcount=19

[pacionet]

I can try .
Ho can I split the PROTHGEN.TMP file in 2 files:

1) 0 -207G
2) 207G-300G

(Windows) ?

[pacionet]

Ok I'll be able to sieve, for the beginning, the range 0-50G for exponents around n=500,000.

As soon as we decide n I'll start with that range.

[MooMoo2]

Why not do 0-207G? From biwema in a previous post in a different thread:

500000
FFT size: 48K
Time 1 test: 500 s
Optimal sieve bits: 57.5
Tests per twin: 23.7M
Twin every G: 90
90% chance after G: 207
CPU years per twin: 376

Originally Posted by pacionet

Ok I'll be able to sieve, for the beginning, the range 0-50G for exponents around n=500,000.

As soon as we decide n I'll start with that range.

Pacionet, I can sieve the rest of the range (50G-207G), if your computer doesn't have enough RAM to sieve the whole 207G at once. It's a lot more efficient to sieve the whole range at once, than to split it into 50G ranges of k and then sieve each k range individually.

Please let me know if you want to sieve the whole range (1-207G) at once, or sieve 1-50G and let me sieve the rest (50G-207G). It's your choice, but the first option is more efficient.