I'm called spirits247 over at amdmb.com and have been debating the dividers question:
Debate over dividers
With the help of some guys over at Alliance forums, I described the theory behind dividers in the above thread and why we should have them at 160 FSB and above (please don't mention chipsets without reading the theory people
🙂 ).
Now I've seen more evidence, including oscilloscope readings, my opinion is - WE HAVE THE 1/5 and 2/5 dividers, and even more
🙂
Here is yet more proof. Using the new 8KHAL00.bin bios (I'm sure this is for the new 8KHAL motherboard, oh well it works!
🙂 ) some people are solving their motherboard reboot issues but also hitting 220+ FSB!
🙂
They are achieving this by setting the RAM to 100 instead of 133, which uses a 3/4 RAM divider on the FSB.
So if you set 220 FSB, your RAM is working at 165 (3/4*220) intead of 220!
🙂
220+ FSB with the new 8KHAL beta bios on an 8KHA+
Since it would be impossible to run AGPs and PCIs at these higher speeds on 1/4 and 1/2 dividers, then I think we can assume the board has the 1/5, 2/5 at 160 and above, and the 1/6, 1/3 at 194 and above, just like the clock gen PDF states.
🙂
REMEMBER, HOWEVER, THAT IT IS NOT THE BIOS THAT DECIDES THE DIVIDERS, WE'VE ALWAYS HAD THEM WHEN PROGRAMMING THE CLOCK GEN TO OPERATE AT CERTAIN SPEEDS (see theory in my other post above).
If the official 8KHA+ bios is as good as the 8KHAL beta one (and why shouldn't it?) then we have one fast motherboard on our hands when RAM tech catches up!
🙂