Elmer, <32 bits x 133mhz is the same as 64x66.> No it is not.
133 is not close to 66, and even 66 is not a current 33MHz bus.
Although formally each effect of frequency on signal integrity
seems to be linear and sometimes goes as a square root, the
increase in frequency and rise/fall times trigggers an avalanche
of technical problems that are expensive to handle.
Mutual inductance, mutual capacitance, overlapped bias currents,
skin/proximity effects, ringing/reflections, ground bounce,
growing emissions .. the list goes on and on.
All this require serious design, need in controlled impedance
traces/vias, both-ends terminations, matched impedance connectors
and IC packages, tighter PCB tolerances and material properties...
That is why the Intel's AGP/PCI-66 "bus" is limited to a single
connector only.
Ali #2
<Ask your nurse, she'll do the math for you, after all,
she's there to help you.> And now shut your stinking mouth
up and never show up your ignorant nose again in technical
matters on this threads. Is that clear now? - Ali #1
(Disclaimer: the latter remark was solely inspired by the
tone of the original poster (Elmer Phud), and is therefore
impedance-matched. For those who are too sensitive to the
reality of "discussions" regarding investor's matters,
feel free to skip this second part of the reply)
|
