lindsay@gandalf.cs.cmu.edu (Donald Lindsay) (11/09/90)
In article <0093F6A8.35049DA0@KING.ENG.UMD.EDU> sysmgr@KING.ENG.UMD.EDU (Doug Mohney) writes: <concerning the 586 rumors> >I suspect you'll see hardwired support for Windows primatives on the chip. What sort of support are we talking about, and how much difference will it make? -- Don D.C.Lindsay
mash@mips.COM (John Mashey) (11/10/90)
In article <11030@pt.cs.cmu.edu> lindsay@gandalf.cs.cmu.edu (Donald Lindsay) writes: >In article <0093F6A8.35049DA0@KING.ENG.UMD.EDU> > sysmgr@KING.ENG.UMD.EDU (Doug Mohney) writes: > <concerning the 586 rumors> >>I suspect you'll see hardwired support for Windows primatives on the chip. > > >What sort of support are we talking about, and how much difference >will it make? This is a worthy instance of a kind of design issue that has often come up: how do you decide what to include in a chip, and specifically, how much (that ought to be) software gets included, and how well does that work? There's been a lot of marketing lately about "putting the human interface on a chip", whatever that means. 1. If somebody says "We're going to build Motif or OpenLook into the chip" I'd run in fear. 2. If somebody says "We're going to build X in" I'd run, although not in quite as much fear. 3. If somebody says "We've got a select set of primitivies which assist the inner loops of 2D (or 3D) operations, and here they are.." Then, I might or might not know whether they'd picked the right primitives, but at least I'd think they had the right approach. (For example: can anybody who KNOWS comment on the i860's graphics operations? Do you use them that way? Are they the right things?) As examples of #1 gone amuck, back in a former life, I got a call from the marketing side of a chip organization, who said "The designers say we should call you for advice. We're designing a new custom CPU, snd we want to build UNIX into it. How should we do that?" my answer: uhh, uhhh. (Just what I need, a CPU with UNIX built-in :-) I asked them if they meant what features would be good for UNIX, and they said, no, they really wanted to embed UNIX itself into the silicon.... Other examples where hardware was designed with very specific languages or other software in mind: 1) Burroughs B5500 (for Algol): I'd call this successful in many ways, as one of the earliest, and most innovative combined hardware-software efforts, and certainly an architectural family that's lasted a long time, even if the specificity of the hardware has sometimes made it more difficult to run as wide a range of programming languages. 2) Intel 432: not so successful. I would claim that most of the most successful products are those that understood carefully the nature of the necessary primitives that could accelerate software in ways difficult for software to duplicate, without getting too specific that it couldn't adapt to later changes in expected software. Maybe people have examples or counter-examples. -- -john mashey DISCLAIMER: <generic disclaimer, I speak for me only, etc> UUCP: mash@mips.com OR {ames,decwrl,prls,pyramid}!mips!mash DDD: 408-524-7015, 524-8253 or (main number) 408-720-1700 USPS: MIPS Computer Systems, 930 E. Arques, Sunnyvale, CA 94086
ingoldsb@ctycal.UUCP (Terry Ingoldsby) (11/21/90)
In article <42992@mips.mips.COM>, mash@mips.COM (John Mashey) writes: ... > This is a worthy instance of a kind of design issue that has often > come up: how do you decide what to include in a chip, > and specifically, how much (that ought to be) software > gets included, and how well does that work? There's been a lot of marketing > lately about "putting the human interface on a chip", whatever that means. ... > Other examples where hardware was designed with very specific > languages or other software in mind: > > 1) Burroughs B5500 (for Algol): I'd call this successful in many ... > 2) Intel 432: not so successful. Aren't you forgetting one of the most successful symbiotic relationships? I speak of VAX/VMS. I took a VMS Internals course a few years ago, and was struck by the frequency with which the VAX architecture just happened to be exactly what was needed for VMS. I finally concluded that if VAX was not designed for VMS, then certainly the hardware architects had a lot of input from the VMS people. Can anyone confirm or negate my suspicions? -- Terry Ingoldsby ctycal!ingoldsb%cpsc.ucalgary.ca Land Information Services or The City of Calgary ...{alberta,ubc-cs,utai}!calgary!ctycal!ingoldsb