Unit is back up and running.
Any insight yet on why the unit's hashrate is so much lower than 830Mhps? It doesn't seem to take long after a restart before it's consistently down in the 500's or 600's. Hopefully it is an issue with cgminer which can be fixed relatively quickly once some more units are in peoples' hands; a fundamental issue with the fpga firmware/hardware could be more difficult to address.
I've been watching and it seems the rates jump around from 600's to 900's. Not sure of the average but other posts have indicated its in the low 800's?
Inaba, how about streaming the desktop of the computer that is hosting the Single? Or point a webcam at it.
- like D&T calculated a few pages back - https://bitcointalksearch.org/topic/m.748164
No mathematical proof here, but if you're mining 1 Diff shares and you mine for at least a few hours, your U: in cgminer will rarely be even 10% out.
On weekdays I mine 18 hours a day (electricity price banding is why - it more than doubles during the 6 hours I don't mine) with 2x6950 ~= 725MH/s which is also in the same general hash rate arena as the BFL and checking 59 logs in the last 2 months where the log was more than an hour, my U: is not once been out by more that +/-10% expected
So that's not mathematically proven, but IMO that number of real figures is good enough for me at least to take notice of the U: value after a couple of hours
( cgminer log file processing script for this you can load in a spreadsheet: http://pastebin.com/JUSNR2tK )