Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!tut.cis.ohio-state.edu!ucbvax!decwrl!decvax!ima!haddock!suitti
From: suitti@haddock.ima.isc.com (Stephen Uitti)
Newsgroups: comp.arch
Subject: Re: quest for breakthroughs (long)
Keywords: architecture, breakthrough, technology
Message-ID: <11780@haddock.ima.isc.com>
Date: 15 Feb 89 22:09:56 GMT
References: <740@tetons.UUCP> <3780@tekig5.PEN.TEK.COM>
Reply-To: suitti@haddock.ima.isc.com (Stephen Uitti)
Organization: Interactive Systems, Boston
Lines: 34

In article <3780@tekig5.PEN.TEK.COM> wayneck@tekig5.PEN.TEK.COM (Wayne Knapp) writes:
=>I'm interested in computer animation.  The biggest problem is
=>memory bandwidth.
=>So what I'd like to see is a two fold solution:
=>    1. RCISC - Register enhanced CISC.
=>       Take something like 680xx, paste in a couple kbytes of
=>       registers that can also hold code.  Use a few of the
=>       unimplemented opcodes to give access to the new memory bank.
=>                                    Wayne Knapp 

Don't invent a new memory store type.  Just put a few pages of
physical RAM on the chip.  This was done for some 8 bit controller
type chips way back when (for different reasons).

The OS can be told to provide access to these special pages to the
programmer, perhaps using existing calls (depending on the OS) to map
them.  The OS may also provide other services which may be handy, such
as real time stuff.

This has the advantages of:
1) no new instructions.  True compatibility.
2) current instruction power is available.
3) no new assembler/compiler tools required.
4) comparatively simple (localized) OS changes.
5) can be expanded dynamically with chip technology.

Intel's separate I/O instructions (for example on the 8080) show how
painful a second address space can be.

Of course, it may be that a peripheral (glorified DMA engine(s))
could do everything you want.  That would only require a "simple
driver"... which you - the programmer could write.

	Stephen.