Re: Super GS

[Rubywand <rubywand@swbell.net>]

Jeff Blakeney writes ...
> 
> On Fri, 14 Apr 2000 01:38:24 GMT, Bill Heineman
> <burger@contrabandent.com> wrote:
> 
> >Even a low-life 66 Mhz 486 can emulate a standard IIgs.
> 
> Would this be using KEGS?  I couldn't get my 486DX4-100 to even do
> stock IIgs speeds with XGS under Windows 95 and even now on my Pentium
> II 233, the fastest speed any of my emulators has reached is about 7
> MHz.
>

     Might be KEGS. According to some 2 year-old benchmarks by Gilles Tschopp
(ref. http://home.swbell.net/rubywand/R026GSEMUS.htm ) KEGS on a 66MHz '486
would come fairly close to standard GS speed.

 
 
> >The real issue is why would you build a machine? Emulators like Bernie are
> >already available and can turn a PowerBook into a nifty IIgs laptop. Unless
> >you REALLY REALLY wanted to plug in your old hard drive cards or other
> >peripherals, there is no compelling reason to build a computer from
> >scratch.
> 
> I want a new machine because I don't want to have to waste hard drive
> space with another operating system, waste time having to wait for
> that OS to boot and waste RAM that the OS needs to run.
> 
> I want a new machine that I can turn on and boot into GS/OS.
> 
> I want a new machine that will run everything faster and access my
> SCSI peripherals faster.
> 
> I want a new machine that takes the great things about the IIgs and
> expands on them without changing the feel of the machine too much.
> Things like higher resolution graphics with more colours and better
> audio.
>

     Right. Basically, like most Apple II users, you do not want to run on a
machine which has to murder its performance to run Apple II software.

     You want to run Apple II software on Apple II hardware. And, very
important, you want there to be a reason to develop more Apple II software and
a reason to continually improve Apple II hardware.



> >On the other hand, doing a PowerPC processor board that plugs into a IIgs
> >slot would be a neat trick.
> 
> The only problem I see with doing this is that you run into the same
> problem that you do with the ZipGSX.  The faster you go, the less
> benefit you get because it spends more time having to slow down to the
> IIgs' 1 MHz bus speed.
>

     Not necessarily. You put everything on the plug-in board-- uP, RAM, ROM,
Video, Sound, serial ports, disk controllers, ... . This lets you plug in a
high-performance IIgs which does not care about bus speed (or RAM/ROM
limitations) of the host GS.

     The catch is that, even if you allow for putting everything on two boards
(and employ top-of-card bus connectors), the level of integration is fairly
high. The card (or card set) would have to be developed and produced by someone
like Apple. There are other problems, too:

o- The computer-on-a-card would not be expandable. This is not a major drawback
if the maker is committed to future upgrade releases-- i.e. you swap in a new
super-card the way PC users swap in new motherboards. Otherwise, it's a big
problem. Either way, you miss the chance to promote user hardware
experimentation and the growth of a New-GS hardware support industry.

o- The computer-on-a-card would be expensive to develop and not especially
cheap to produce. Since you 'get the whole thing' in one chunk, the price is
relatively high. Getting users to invest in the new card could be difficult and
you might never develop a decent user base.



> I've been thinking of a modular design where the processor would be a
> plug in module so it could come in different designs offering
> different speeds and features.  You could have a 65816 module or a
> StrongARM module with ROM code to make it emulate a 65816 or whatever.

     This could work if we are talking about a separate expansion box. That is,
you have a reasonably simple motherboard inside a PC tower case which connects
to your current Apple II (for 'getting started' keyboard, disk I/O, ...).  The
new motherboard could have a socket to permit plugging in any sort of processor
module which acts like a 65C816 (or, in time, a 65C1632).
 
> You could also create an emulated 65832 this way even though the real
> thing never existed.
 ....

     Except, maybe, as a prototype at WDC.

     One nice thing about a plug-in processor module on a simple motherboard is
that you can start at the low end with something simple, like the current
65C816. It's the kind of project any number of Apple II clubs and individuals
might get into. Running at 20MHz with no bus slowdowns and accessing 16MB
superbanks, it would be a nice improvement over the current GS!



Rubywand