Are Conroe mem divider function dependent on CPU multi?

[lorkp]

Athlon64 memory dividers are funky. A divider isn't always as it appears because it depends on the CPU multiplier. For example, a 228 with an 11 multiplier and 200:183 divider yields 193MHZ memory speed. One would think it would be more.

Will dividers with Conroe be the same way, or will they be straight forward?

 

[gobucks]

as far as i can see, it's simply based on a multiple of the FSB, and it doesn't have anything to do with the multiplier.

 

[Suspicious-Teach8788]

A64 does not use memory dividers. You are completely mistaken. For someone who was raised through the FSB era, it's hard to understand, but memory speed is pure memory speed. Yes it's derived from the CPU speed divided by a certain number, but that number is calculated based on a ratio involving the multiplier and HTT. In the end it is not a really a divider.

Conroe works like P4, so they are dividers of the FSB. You operate at a certain ratio like 4/5, 2/3, 5/4, whatever of hte FSB, so it's simple to grasp. With A64s, you have to do insane calculations, and I prefer to do it with a handy tool people create (especially with DFI's odd divisor correct thing now... ugh)

 

[lorkp]

that's good to know, thanks!