Path: utzoo!utgpu!watserv1!watmath!att!dptg!ulysses!andante!princeton!udel!wuarchive!usc!jarthur!nntp-server.caltech.edu!toddpw From: toddpw@nntp-server.caltech.edu (Todd P. Whitesel) Newsgroups: comp.sys.apple2 Subject: Re: New Apple II generation Message-ID: <1990Nov21.050507.26700@nntp-server.caltech.edu> Date: 21 Nov 90 05:05:07 GMT References: <9011210256.AA09328@apple.com> Organization: California Institute of Technology, Pasadena Lines: 64 $CSD211@LSUVM.BITNET (Mark Orr) writes: >work building a RISC replacement for the Mac would lead anyone to conclude >that the Apple II is the last thing on their minds. yep, and that's part of the problem. Apple's spending $40 million fighting the perception that Macs are expensive... >|I'd look into licensing technology from Zip and maybe CVT, and especially the >|ASIC Technologies guys -- you will also need a VLSI ASIC service and a system >|to design the custom chips on. This can cost $$$ rapidly. >Oh, great! You mean we have to fight Apple and Zip too??? No No No. I mean we ought to try to get their help! >As long as it supports 4 Mb SIMMS (and above), I have no quarrels Precisely. >Yeah, I really like the way Todd described it in the //f paper (i.e. mapping >Ensoniq registers to the address space of the '816. How much trouble would >it be to add a DSP (i.e. Motorola 96002, TI 32020, etc.)? Thanks for the compliment, after trying to program the DOC myself it became pretty obvious that memory mapping the DOC RAM and registers would vastly simplify a number of things (and that's not all, heh heh). Putting a DSP on the motherboard is asking a bit much though, but I expected the Sound Direct Slot to take care of those concerns. >Whoa there is right! There's alot we don't know about the ASIC (i.e. How >compatible, how reliable, how much, how many). The guy was talking about I'm not worried about ASIC. I just don't know how long they'll take. >that runs over 20 MHz with an onboard cache. The AMD 29000 is a very fast >RISC chip that even without a cache of any kind could emulate an '816 faster >than the ASIC. I'm sorry, but I don't think so. Let me see some code. I sincerely doubt that a 20 mhz 29K can emulate a 20 mhz 65816, even given the cycle differences. >a Mac LC". Because with these chips (I favor the ARM - it's cheaper) one >could build a system that is competitive in price with the Mac LC, but far >superior in terms of capability. Prices for complete systems would range >from $1500 bare bones to $2500 loaded-to-the-gills, and could have slots and >built in language and toolbox support...i.e. a 32-bit BASIC with full access >to system resources and the equivalent of MacApp in ROM. Further it could >(and should) embrace Unix and X-Windows (something I seriously doubt that >the '816 could ever do). Sounds extremely seductive, and it's certainly the right direction, but think about the consistency of it: you might as well buy an ARM and write a GS emulator. Still wouldn't be as fast as the machine whose creation we're pondering -- when it was emulating. Your point that it would capable of a lot more than emulating a GS is well taken, but I feel that once the decision to use a different CPU has been made, GS emulation becomes less and less worth the effort. I still think that a better system can be had by fixing Apple's mistakes rather than by building a whole new computer and forcing it to emulate a GS. Todd Whitesel toddpw @ tybalt.caltech.edu