That was to be expected [for your task]. The first 2 did not agree, so a 3rd had to be the judge and jury which of the first 2 is OK. It would have been worrisome of a _3 would have been needed i.e. the 3rd also turning PVer. The validation rules on the server are much more extended than are integrated in the application i.e. [sadly], logs are not all-telling... slightest of hope there was that W10 was wonky, but production reports have dismissed that opinion. If logs 1 on 1 compare the only thing one can conclude is that neither were interrupted during the run [resumes without LAIM do create deviating output files].

Only had 1 of the 64 bit make. It went valid without need of a 3rd opinion.

BETA_HST1_000001_000198_AC0004_T000_F00005_S00001b_1 121 beta22 Valid 50.481 2.3554 0:02:21:20 2.3783 0 59.86 25.41 99.04%

Thought it was the fastest T000 on the rig so far, but then remembered that the day before upped the cycles by 200MHz :D)

Great news!