This is very bizarre. Got it working, but very bizarre.
After a few BIOS resets (as it would no longer post - and managing to corrupt a Windows installation along the way), it allowed me to POST the chip in the 100MHz setting even though it should be 133MHz. From there I could "overclock" it to 133 (well, beyond too but how stable's another question). . . so it seems normal now. A bit toasty, but got it working
Begs the question why would you have a 133/166 setting if you cant use it - Gigabyte implicitly states that this setting is for palamino/t-bred//bartons etc. (any XP chip), yet you use the 100 setting meant for early Socket A Athlon/Durons.