Until recently, I hadn't had the spare time to experiment. Now that I have, the situation is even more confusing... Or perhaps I should say mysterious.
As I said, I was able to use approximately the method here,
http://kmwoley.com/blog/?p=345, to get the 64MB Flash drive to boot. I was also able to get the 2GB Flash drive to boot using almost the same method. I stopped short of copying Microsoft's Boot Manager in the 2GB drive's case. The interesting thing is that now, no matter what method I use (Linux Live USB Creator, UBCD, UNetbbotin, or the Universal USB Installer), both Flash drives boot. I haven't been able to test if this is an F3 specific issue, yet. Now that both drives I've been using for testing boot regardless of the method used to prepare them; I'm unable to determine if the problem is revision specific.
Another interesting wrinkle that I didn't notice until after making the Flash drives bootable (with DISKPART) is that both Flash drives appear as '"
H"ard "
D"isk "
D"rive(s)' from within Q-Flash as opposed to "Floppy" drives. Previously I chalked this up Q-Flash employing floppy emulation (i.e. CD boot disks). Especially since Q-Flash correctly reported the capacities of both drives even as 'floppies'.
From these experiences, I've come to the conclusion that the GA-785GMT-USB3's BIOS is not picky about the bootloader per se. But the BIOS
is picky about the active status on the Flash drives and/or the partition table when using any other method to create them than Microsoft's DISKPART. Right now, I suspect it's the partition table (or lack thereof) that is the key to this dilemma. I guess my next step is to wipe the partition table on one or the other drive and see if that has any affect. If that fails to 'reset' the drives to a non-bootable state, I'll have to scrounge up some more test drives.
I may leave this topic go dormant for a while... Figuring out this puzzle has lost some of its urgency, if you know what I mean?
Nevertheless, I'll be back with my findings eventually.
