I have a spare machine (a test server but mostly idle) with the aforementioned 400MHz Pentium II and 256Mb RAM. The deadline shown for report is April 2008 but the estimated time is just over 2 years! I noticed the recommendations are for 1.6GHz processors.

Should I put my spare cycles to better use on some other project?

It\'s not worth it, especially with so little ram. These are supercomputer models, and need a computer with a fair amount of power. Anything less than the minimum recommended, and the models are likely to keep crashing.

You\'re better off using it for one of the other projects.

Hi Skymist

I\'d say that two years is too long, and you might be disappointed to find that some of your results arrived in Oxford too late to be used by the scientists, or you were advised to abort the model before the end. Another problem is that each model is supposed to have 512Mb of RAM. If the machine was only being used for the model and the graphics weren\'t used, 256Mb might just be enough, but it\'s still risky.

You could keep an eye on the cpdn news (see my sig) for announcements about launches of different, and probably shorter, models. But you\'ll need to look at the RAM requirements because the rumour we hear from Oxford is that future models are likely to require as much or more RAM, not less.

So yes, I\'d say there are other projects that would use this machine more usefully for science and more satisfyingly for you.

Edit - Les was quicker, and I\'m relieved to see that his recommendation is the same.

---

Cpdn news

There is a list of the different projects, there recommendations and crunchingtimes (measured at a 2 GHz system) at the BOINC wiki : http://boinc-wiki.ath.cx/index.php?title=Choosing_a_BOINC_Powered_Project

And the same for the alpha and beta projects : http://boinc-wiki.ath.cx/index.php?title=The_Current_List_of_Alpha_Test/Beta_Test_Projects

---

It\'s not worth it. It wouldn\'t finish the work unit by the due date.