Path: utzoo!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!samsung!zaphod.mps.ohio-state.edu!tut.cis.ohio-state.edu!ucbvax!decwrl!polyslo!vlsi3b15!batman!nicholaA From: nicholaA@batman.moravian.EDU (Andy Nicholas) Newsgroups: comp.sys.apple Subject: Re: 20 Mhz 65816 Summary: Gass Attack? Message-ID: <1159@batman.moravian.EDU> Date: 2 Mar 90 17:42:43 GMT References: <10393.infoapple.net@pro-generic> <1990Mar2.081825.28966@spectre.ccsf.caltech.edu> Organization: Moravian College, Bethlehem, PA Lines: 60 In article <1990Mar2.081825.28966@spectre.ccsf.caltech.edu>, toddpw@tybalt.caltech.edu (Todd P. Whitesel) writes: >>No, it's NOT microcoded. Inherently, that's one of Mensch's problems... >>everything on the 65816 is done at the gate level. He hard-wires EVERYTHING. > > I'll make what I will of the 65816 data sheet: > > "The Timing Control Unit keeps track of each instruction cycle as it is > executed. The TCU is set to zero each time an instruction fetch is executed, > and is advanced at the beginning of each cycle for as many cycles as is > required to complete the instruction. Each data transfer between registers > Timing Control Unit." > > If that isn't microcode then PLEASE TELL ME WHAT IS!! Actually, it sounds like the control logic for a 1984 vintage microprocessor. Todd, you get excited easily, don't you? Maybe we don't have the same definition of microcode. I was taught that the 68000 was microcoded. This came out of an Andrew S. Tannenbaum book (2nd Ed, not the newly released 3rd). Read the chapter on the MIC and MAC microprocessors. 2 years ago, I had to write a simulator in C which simulated the flow of the microcode for the MIC machine, so I'm taking a wild hunch that I know what I'm talking about. If the 65816 were really microcoded (still haven't convinced me, according to what I've been exposed to), then wouldn't it have a microsequencer and a microcode store on the chip? Look at the mask sometime, I don't think you'll be able to find anything like that... the ASIC guys are *NOT* deriving their speed increase from going from microcoded to gate-array. They are deriving their speed increase by designing their chip more efficiently with a decent CAD system to predict defect-free silicon. A human did the mask layout for WDC. A computer did it for ASIC, and it looks like it worked. Call the ASIC guys and ask them if the WDC 65816 is microcoded, or call WDC and ask if it's microcoded. C'mon guy, if you really want to know, don't post this stuff on the net, give the folks that KNOW a call... :-) >Looks like Mensch is just suffering from the price of cutting corners. He needs >to either redesign his mask so it will be easily manufacturable or maybe he >should license the ASIC guys' mask. (I'd like to see him cooperate with them >to make the 65816 a reality, but I don't know if any of them would go for the >idea...) If you look at the transscipts from America Online, you'll find out that the ASIC guys have had very few words from Bill Mensch, and the ones that were said were not too nice. Mensch didn't think they could do it, and since it steals some of his business (of which the Apple II is a TOTALLY INSIGNIFICANT AMOUNT), I doubt that ASIC and WDC will get along. Bill is proud because the 65c02 (WDC's) is something like one of the only cpu's certified by the FDA for implantation into the human body. WDC gets a LOT of their revenues from medical stuff... > Todd Whitesel andy -- Yeah!