Nope, they are dual booted between Windows and Linux, so Linux is running AK optimized, and I moved some of them over to standard seti binaries in Windows. Hopefully they will all be up and running in a short while.

---

Checkout http://seti.smlug.net for optimized Sun binaries.

Well after some time I.ve managed to upload from other host same result.

Wandering willie host Q6600

1093228666 4052340 18 Dec 2008 22:04:31 UTC 19 Dec 2008 18:29:21 UTC Over Success Done 249.51 62.08 45.44
1093228667 4608661 18 Dec 2008 22:04:33 UTC 19 Dec 2008 8:21:31 UTC Over Success Done 25,346.59 45.44 45.44

Michael host. Q6600

1091890751 4687023 17 Dec 2008 21:38:45 UTC 18 Dec 2008 14:24:29 UTC Over Success Done 18,773.01 46.33 46.33
1091890752 4491788 17 Dec 2008 21:38:46 UTC 19 Dec 2008 10:18:12 UTC Over Success Done 253.20 59.77 46.33

I have checked with Beta results most were single replication.

So should we be doing the same. Or running 6.05 CUDA with 6.05 CUDA and not against MB.

Michael

...
2) Given the normal mechanism for credit, it doesn't seem reasonable on its face to assign credit for a WU based on a cuda result to a non-cuda machine - the actual cobblestones expended are different.

The credit method here is specifically implemented so that hosts doing the same work should report the same fpops_cumulative value used by the Scheduler to calculate credit claims. The stock CPU apps 6.03 and earlier in effect "count" flops based on the most generic code even though vectorized code may be in use for some functions. Optimized apps use the same counting. The 6.05 CUDA app ought to do the same, but obviously is getting extra "counts" for some WUs. I expect that's a temporary flaw which will be corrected after more serious problems have been fixed.

                                                               Joe

...
So should we be doing the same. Or running 6.05 CUDA with 6.05 CUDA and not against MB.

Michael

6.05 CUDA MB definitely ought to produce the same results as 6.03 and earlier non-CUDA MB, if not it needs to be fixed. Having both running here and no Adaptive Replication makes quickly finding any differences possible.

                                                                Joe

I seem to be consistantly underclaiming slightly on my results on this pc. I've done mostly shorties, show the difference hasn't been great yet, but as soon as I start getting some of the longer ones done, I expect the difference probably will increase.

---

I seem to be consistantly underclaiming slightly on my results on this pc. I've done mostly shorties, show the difference hasn't been great yet, but as soon as I start getting some of the longer ones done, I expect the difference probably will increase.

I'm experiencing similar for shorties. Underclaiming slightly...usually less than 1/2 a CS.

Task list for computer 4715682.

For longer WUs it's usually an Overclaim, not an underclaim.

Yep, I'm overclaiming on a longer task now too. At least on this one.

---