Continuous failures is bad for the project and bad for the cruncher.
For example,
Tasks from Computer (GeForce 9800 GTX0):
5199396 3318610 3 Apr 2012 | 6:46:22 UTC 3 Apr 2012 | 7:12:32 UTC Error while computing 789.86 18.19 --- ACEMD2: GPU molecular dynamics v6.16 (cuda31)
5199359 3319868 3 Apr 2012 | 5:35:16 UTC 3 Apr 2012 | 5:46:46 UTC Error while computing 79.96 1.42 --- ACEMD2: GPU molecular dynamics v6.16 (cuda31)
5199260 3304669 3 Apr 2012 | 6:46:22 UTC 3 Apr 2012 | 7:12:32 UTC Error while computing 105.21 2.62 --- ACEMD2: GPU molecular dynamics v6.16 (cuda31)
5199259 3319810 3 Apr 2012 | 5:35:16 UTC 3 Apr 2012 | 5:42:16 UTC Error while computing 123.96 1.81 --- ACEMD2: GPU molecular dynamics v6.16 (cuda31)
5199217 3319791 3 Apr 2012 | 5:35:16 UTC 3 Apr 2012 | 5:53:58 UTC Error while computing 892.56 15.23 --- ACEMD2: GPU molecular dynamics v6.16 (cuda31)
5198388 3314156 3 Apr 2012 | 5:35:16 UTC 3 Apr 2012 | 5:46:46 UTC Error while computing 69.74 1.36 --- ACEMD2: GPU molecular dynamics v6.16 (cuda31)
5185962 3311487 31 Mar 2012 | 9:38:50 UTC 31 Mar 2012 | 9:48:14 UTC Error while computing 167.70 3.68 --- ACEMD2: GPU molecular dynamics v6.16 (cuda31)
5185887 3311439 31 Mar 2012 | 9:38:50 UTC 31 Mar 2012 | 9:52:26 UTC Error while computing 57.72 1.58 --- ACEMD2: GPU molecular dynamics v6.16 (cuda31)
...

---

FAQ's

HOW TO:
- Opt out of Beta Tests
- Ask for Help

Some projects (or some WUs) are running some not, but i deactivaded my 9800 and so on cards for gpugrid cos i dont want to produce failures while it burns energy (for nothing then).. Sadly cos such card could get round 30000 credits per day on short units when not failing some WUs.

---

DSKAG Austria: http://www.dskag.at