I think it's gonna be a long, long time

[panic]

Hey guys,

It's my first time on the forum, and i'm glad to be part of GIMPS I've been going for about 2 weeks and I used to a few years back aswell.

I've very recently (about 5 minutes ago) got a Pentium III @ 500MHz doing LL (100) "first-time" tests, and although already having a vague knowledge of how incredibly long it takes to do them, I was shocked to be told that it would take 626 days!

I don't know whether it's worth the time and effort to let it carry on working, since it said that the exponent might be re-assigned after a year anyway. Any advice?

Cheers

[Uncwilly]

My advice is to change the work type for that machine to TF-LMH.

Trial Factoring assignments are much shorter in duration. And those in the higher ranges (commonly known as LMH) are even faster and more well suited to your machine.

[cheesehead]

Originally Posted by panic

I was shocked to be told that it would take 626 days!

I don't know whether it's worth the time and effort to let it carry on working, since it said that the exponent might be re-assigned after a year anyway. Any advice?

Be sure to send in regular progress reports. Exponents likely to be reassigned are those showing no recent progress.

[Uncwilly]

Originally Posted by cheesehead

Be sure to send in regular progress reports. Exponents likely to be reassigned are those showing no recent progress.

And perform a back-up of the directory about once a month.

Also, a double check would be faster, as it is a smaller number.

[panic]

Originally Posted by Uncwilly

My advice is to change the work type for that machine to TF-LMH.

Trial Factoring assignments are much shorter in duration. And those in the higher ranges (commonly known as LMH) are even faster and more well suited to your machine.

Ah yeah thanks, i'm not prepared to wait nearly 2 years for a result! A lot of Elton John to listen to if I let it carry on!

Cheers

[pegaso56]

Hi. May be a good idea to change to Linux, use mprime instead of prime95 in order to optimize cpu resources, and if you have enough memory you could try doing p-1.
Regards, Carlos

[panic]

Originally Posted by pegaso56

Hi. May be a good idea to change to Linux, use mprime instead of prime95 in order to optimize cpu resources, and if you have enough memory you could try doing p-1.
Regards, Carlos

Coincidentally I already was using linux 32-bit mprime with nothing else running I've now assigned P-1 factoring.

Cheers

[jasong]

I forget how much power those P3s use, but purely from an electrical cost perspective you might do pretty well with the P3. There was someone a good while back who put a bunch of P3 stuff together, about a kilowatt of power draw, I think. That person did really well from a total cost perspective and it might still be true for P3s.

You can do whatever you want, I'm just killing time with this post. :)

[henryzz]

Originally Posted by jasong

I forget how much power those P3s use, but purely from an electrical cost perspective you might do pretty well with the P3. There was someone a good while back who put a bunch of P3 stuff together, about a kilowatt of power draw, I think. That person did really well from a total cost perspective and it might still be true for P3s.

You can do whatever you want, I'm just killing time with this post. :)

i would be surprised if a kilowatt-worth of p3s would outdo one cheap Q6600
plus a Q6600 would use maybe a quarter of the power and possibly cost less than getting a lot of p3 parts together

[Reed_Young]

Originally Posted by jasong

I forget how much power those P3s use, but purely from an electrical cost perspective you might do pretty well with the P3. There was someone a good while back who put a bunch of P3 stuff together, about a kilowatt of power draw, I think. That person did really well from a total cost perspective and it might still be true for P3s.

You can do whatever you want, I'm just killing time with this post. :)

Combined with another suggestion I read earlier today, booting Linux without a hard drive using NFS and PXE, I can believe somebody got enough P3 processors to do more with a kilowatt than with one bleeding-edge standalone system that also requires one kilowatt. If you can find that story about the P3s I'd read it.

Reed Young