Before the 22nd April I believe I was producing almost 100% validated results. Since we entered April 23rd, almost all my work is erroring. Although just now the majority of the work, when running for about 335 seconds or 600+ seconds is validated quickly. The erroring WUs seem to run for about 5 seconds then fail.

The set up for the OS and GPU are - win XP pro SP3, and an ATI HD3850 on an AGP slot. The ATI driver is AMD's 11.6 with AGP Hotfix.

I have looked at the WUs names and seem, in the majority to get the following separators -

ps_p_82_1s_dr8_4
and
de_p_82_1s_dr8_4

and a number of _dr8_8 work

The ones that error seem to be similar.

Looking at the completed work unit details, I see all are giving errors on loading -

Error loading Lua script 'astronomy parameters txt'
and
Error reading astronomy parameters from file 'astronomy parameters txt'

Anyone able to give an opinion as to what is going on to cause the errors, and what, if anything, I can do to return almost 100% valid results again?

I have left my computers unhidden, and to one in question is the old dual Xeon with a Nvidia 9500GT as well as the ATI HD3850. The latter contributes a small part to PrimeGrid.

---

Go away, I was asleep

The real error is a bit more down in the logs.

The gpu app for your HD38xx is failing.
It looks very similar to what I see on my own HD5850 with too other runs.

See http://milkyway.cs.rpi.edu/milkyway/forum_thread.php?id=3205

Looks like while they are playing with the parameters for the new runs they are sometimes getting to close to the limit and that is causing the runs to fail on some system but still run on others (possibly older gpus vs. newer gpus).

The problem seems to sorted itself, and I am back to 100% error free crunching ...

---

Go away, I was asleep

Check out the recent posts in this thread:

http://milkyway.cs.rpi.edu/milkyway/forum_thread.php?id=3205