Changed...still nothing found. Thanks again. :)

---

You will be assimilated...bunghole!

hopefull you have no crashes like me

---

you should modify it to:
findstr /M "<true_angle_range>0.0[0-9]" "C:\Documents and Settings\All Users\Application Data\BOINC\projects\setiathome.berkeley.edu\*.*" > con:

that will show you tasks with range from 0.00x to 0.09x
your version shows only 0.00x

That seems like it's getting beyond the VLAR range

findstr /M "<true_angle_range>0.0[0-1]" "C:\Documents and Settings\All Users\Application Data\BOINC\projects\setiathome.berkeley.edu\*.*" > con:

checks and reports them up to 0.1 from what I've been getting.

---

Yes.
To be shure to have no low angel range ones set to
findstr /M "<true_angle_range>0.[0-1]" "C:\Documents and Settings\All Users\Application Data\BOINC\projects\setiathome.berkeley.edu\*.*" > con:

---

I will be suspending GPU and using CPU until the next version of the CUDA client comes out. Performance is good and I've had no crashes yet, but a second workunit locked up for no reason and needed to be manually suspended / resumed for computation to keep going. CUDA seemingly isn't quite at the point where it can be left to run without babysitting. :)

---

You will be assimilated...bunghole!

next version is out, and same problem.

---

It's not just slow graphics cards, because I get the same issue with my GeForce 260. It usually happens when the work unit hits 17 seconds of CPU time (and is still at 0%). Most of the time the percentage will go up after this lag period. Sometimes the unit will just finish with a Computation Error, and there is also a very slight chance of it freezing up the video card completely, resulting in Windows going to crap resolution and saying the video card had a problem and needs to restart to restore it. On restart, the problem shows up as: Error Message: STOP 0x000000EA THREAD_STUCK_IN_DEVICE_DRIVER (Q293078).

---