My 3000+ amd athlon xp has very few computing errors..but yet my 4400+ it seems about half are computing errors..but yet when it passes it back out the other person finish's it...why is this?

---

also Theres an option to tell it how to run for X amount of time...shouldnt it just run the WU till its over? and not stop at a specific time? whats the difference between 2 hours and 8 hours...is it more beneficial to set it higher or lower for the project?

---

My 3000+ amd athlon xp has very few computing errors..but yet my 4400+ it seems about half are computing errors..but yet when it passes it back out the other person finish's it...why is this?

This FAQ might answer some of your questions about the Time setting and who it affect the work unit runs. As to the basic problem you are having, how do you have your Keep in memory setting set, and also what run time are you using.

---

Moderator9
ROSETTA@home FAQ
Moderator Contact

Elsewhere, there's a thread about selecting the iterations that are being run (options on an earlier client). The project needs about 10,000 iterations or models run; and it doesn't matter if it's 1000 machines running 10 models, 100 machines running 100 models, or 10 machines running 1000 models, or any combination that adds up to 10,000. So for the science - it doesn't matter what your max cpu time setting is; it'll pass out the WU enough so that it gets its 10,000 models.

If making the change to always leave the client in memory doesn't get rid of the 50% failure rate, there was a request by the project staff to change the max cpu time setting down to 2 hours to see if that reduced the error rate. And if you continue to get errors, they'd appreciate if you'd sign up for the Rosetta Alpha project so they can help identify why the client is failing so often on one of your machine and not others.

---