I mean, I have 7 failed WUs in the last 2-3 weeks: 1 Ibuch, 1 Gianni, 5 Toni.

The Ibuch and Gianni ones were completed succesfully by the second user, the Toni sometimes by the fifth.

I have a stock frequencies 8800GT with aftermarket cooler which runs at really low temperatures, WindowsXP, a slightly OCed Athlon64 X2 and enough RAM for everything.

Driver 191.07, BOINC 6.6.38.

I am looking into the WUs. However, my feeling is that the higher failure number is due to the fact there is a larger number of this kind of WUs.

Another thing to see is whether all of the HERG WUs fail more, or only the D300s. In principle, they should be all very similar (and very similar to the other targets, as well).

For the followup, please see here

the WUs I was crunching in December look like this.

Failures

Toni HERG 5 (4 of them D300s)

Success

Gianni 5
Ibuch 2
Kashif 2

That makes a total of 100% errors for the Toni-WUs on my PC. Please fix this soon...

the next TONI-WU just failed...that makes six errors out of 6 TONI-WUs this month. I guess, I will abort all new TONIs until this problem is solved...

I have also noticed the TONI WU's failing frequently (all the time). Being a noob with GPU type grid computing, I wnated to observe a bit before posting. My findings are (given as only the most recent past 30 WU's are available for review): 4 of 4 TONI WU errors (2 HERGD and 2 HERG) and one IBUCH failure. To the best of my recollection, I have not had a single failure with KASHIF_HIVPR, GIANNI_BIND or OTTO_HERG WU's to date. 5 errored WU's out of 30 processed seems really excessive and especially annoying given these WU's usually wait to fail after 8 to 12 hours into the process. I also have another TONI in cache and am thinking of aborting instead of wasting another 8-12 hours time unless given a reason why not.

I am using a GeForce GTS 250 1GB (MSI, not the EVGA OC type) at stock clocks with plenty of disk and memory space available to the project. Projects are WCG using CPU and GPUGrid using GPU (800/200 resource values respectively). Grid crunching since 2/2003.

I am looking into the WUs. However, my feeling is that the higher failure number is due to the fact there is a larger number of this kind of WUs.

Another thing to see is whether all of the HERG WUs fail more, or only the D300s. In principle, they should be all very similar (and very similar to the other targets, as well).

For the followup, please see here

The ones failing on my PC are all Toni Herg Wus.

I'm fed up with aborting all Toni WUs on my PC, could you please fix this?

I have to agree. If you cannot say what's wrong with the Tonis, please remove them. I have even underclocked my card, and still got an error. Bad coding like this makes me less inclined to donate computer time.

Sorry for the inconvenience caused. This has been discussed at length in a relevant thread.