but I am curious how he can achieve the level of performance he initially claimed

http://www.8052.com/forum/read.phtml?id=131602

gave you all you need, if you want to argue 'possible' why did you not do it there where I proved it possible instead of waiting a few posts and then blabbering again?

I'd even be interested in how he persuades the internal UART/timer combination to operate at that rate when he's clocked to produce 100 MHz as his system clock.
http://www.silabs.com/public/do...2x-13x.pdf

BTW I sted "rougconvert' GOT IT! it is 96MHZ since USB requires a multiple of 12

I've found it very tricky getting a 30 MHz MCU at 30 MHz to bit bang a 3 Mbps rate.
WHO ON EARTH suggested a comaprison with a 12 clocker (the same generation as the 8255)?

That's why I was fascinated by Erik's reference to a speed switch, which, of course, was a misnomer, since his MCU simply stretches cycles, as does mine.
misnomer???????? does "stretches cycles" not change the execution speed or is that another of your fantasies about the elegance of using slowpokes?

since he's stated many times that he hasn't ever used Mode 0,
Yes I have stated that
and his MCU's Mode 0 timing won't go that fast.
how do you know that?

What kind of a display requires high-rate refresh, uses serial interface, yet doesn't require some sort of rigorous protocol?
go figure

The high rate is probably for dot-wise refresh, but he'd be the one to ask.

You keep diiging for me to realease proprietary information. Your wild guesses clearly show the importance of keeping it propritary (this not ment as a criticism of you, but the fact is that if the proprietary was easy to guess then why keep it proprietary)

Erik