expire times?

[shifted]

I love finding small projects with few users, so my paltry assemblage of computers are enough to actually count. How long before work units expire? I have the client running on a 486 (33mhz) among things, because i hate wasting cycles, and i want to know if it's worth running the sob client on it. (The other systems range from a couple athlons to a p166mx)

[Firebirth]

If your client don't submit a block within two weeks; the number will fall out of your hands...

If you don't think your client can make that... consider finding version 0.9.8 (i don't know where!) instead of 0.9.9, since it takes 2.5 times smaller blocks than 0.9.9! And I don't guess that the performance increase in the latest version would affect your 486 anyhow...

Tarje

[jjjjL]

my atlon can complete a block in an hour. admittedly, a 486 will not do that but it may be able to do it in less than two weeks.

honestly i don't know if it will even run on a 486. if you get it working let me know.

-Louie

[smh]

The GIMPS client is about 2000 to 2500 times faster on a AthlonXP compared to a 486(33). Even a P60 is 50 times faster.

Assuming this will be about the same for the SoB client:

If it takes an AthlonXP one hour then i doubt the 486 will do it within 2 weeks. Even if it did in two weeks to do one block, then it would take 6 years to do a current test!!

It might be wise to wait a day or two until the lower N's for the next K will be available for testing. If you can get a low N then you can complete a test in a month or two. But i seriousely doubt you can do one block in two weeks since the cEMs will be lower for lowers N's so a block would take even longer

[shifted]

Runs perfectly fine on a 486 with 8 mb ram. Linux is good on small things

In fact, i used to use the same machine as a web/email server, and it handled a moderate load just fine (sufficient to fill a T1 with static html). Though it helped to custom compile apache (or whatever) with few extensions so it used less ram.

Granted, the sb client does use over 16 mb or ram, but it appears that it accesses memory in close regions for the most part, so swapping is minimal.

[shifted]

Originally posted by smh
The GIMPS client is about 2000 to 2500 times faster on a AthlonXP compared to a 486(33). Even a P60 is 50 times faster.

Assuming this will be about the same for the SoB client:

If it takes an AthlonXP one hour then i doubt the 486 will do it within 2 weeks. Even if it did in two weeks to do one block, then it would take 6 years to do a current test!!

It might be wise to wait a day or two until the lower N's for the next K will be available for testing. If you can get a low N then you can complete a test in a month or two. But i seriousely doubt you can do one block in two weeks since the cEMs will be lower for lowers N's so a block would take even longer

Thanks for the info... i guess i'll have to find some other use for that obsolete machine. I had it running eccp109 before, and it did actually generate results... usually one a day.

Just so we don't have to wait two weeks, if anyone can manually fix the number pool, it was working on 2868009.

[smh]

Originally posted by shifted


Thanks for the info... i guess i'll have to find some other use for that obsolete machine. I had it running eccp109 before, and it did actually generate results... usually one a day.

Just so we don't have to wait two weeks, if anyone can manually fix the number pool, it was working on 2868009.

I was just making a wild guess, only you can tell how long it would take to complete a block. Didn't wanna keep you from running the client, just wanted to warn you that it could take quite long to complete a test