torrie@cs.stanford.edu (Evan Torrie) (06/10/91)
melling@cs.psu.edu (Michael D Mellinger) writes: >blitter to your computer). The 50MHz 040 is due in 1Q 92. The >article also says that the 68050 will be available sometime next year. >It will double the performance of the 040 at the same clock speed. Actually, reading between the lines, this announcement sounds like the beginning of the end for the 680x0 line. The article says that the '050 will be an evolutionary upgrade, using just improved process fabrication and a few small architectural changes to get its increased speed. All in all, it sounds very much like the situation with the 68030 vs 68020 - claimed "double the performance" with an evolutionary upgrade, which actually ended up with only 25-30% improvement. Why do I see this as the beginning of the end? Well, by all accounts, Intel's 80586 will be as revolutionary a chip from the 486 as the 386 was from the 286. Reports of an i860 on board, superscalar processing etc seem to dominate any discussion of the 586. If Motorola is just planning an evolutionary upgrade for the 050, I can't see how this is going to compete head-to-head with the 586. My guess is that Motorola has been told by its major 68030/040 users that they're ready to switch away from the 680x0 family [e.g. NeXT, Apple with possibly the 88K, HP with their own PA RISC] for their high-end products, and so Motorola is not putting too much effort into designing high-end follow-ups to the 68040. What do you think? Will the 68050 be the last high-end member of the 680x0 line? -- ------------------------------------------------------------------------------ Evan Torrie. Stanford University, Class of 199? torrie@cs.stanford.edu "If it weren't for your gumboots, where would you be? You'd be in the hospital, or in-firm-ary..." F. Dagg
rjc@wookumz.gnu.ai.mit.edu (Ray Cromwell) (06/10/91)
In article <1991Jun10.072945.8821@neon.Stanford.EDU> torrie@cs.stanford.edu (Evan Torrie) writes: >melling@cs.psu.edu (Michael D Mellinger) writes: > >>blitter to your computer). The 50MHz 040 is due in 1Q 92. The >>article also says that the 68050 will be available sometime next year. >>It will double the performance of the 040 at the same clock speed. > > Actually, reading between the lines, this announcement sounds like >the beginning of the end for the 680x0 line. The article says that >the '050 will be an evolutionary upgrade, using just improved process >fabrication and a few small architectural changes to get its >increased speed. All in all, it sounds very much like the situation >with the 68030 vs 68020 - claimed "double the performance" with an >evolutionary upgrade, which actually ended up with only 25-30% >improvement. The 68040 is 300% faster than the 030, I wouldn't start doomsaying yet. > Why do I see this as the beginning of the end? Well, by all accounts, >Intel's 80586 will be as revolutionary a chip from the 486 as the 386 >was from the 286. Reports of an i860 on board, superscalar processing >etc seem to dominate any discussion of the 586. > > If Motorola is just planning an evolutionary upgrade for the 050, I >can't see how this is going to compete head-to-head with the 586. Let's not engage in speculation of what the 050 will or will not be based on a blurb. I don't know what Motorola is planning and I think it's bad form to compare 2 chips before either of them have been officially announced. The 040 just started shipping not too long ago, so the 050 if it is to come out in only a year from now, will probably be an evolutionary upgrade like you said, however, if Intel is planning such drastic architecture changes the 586 may take a lot longer and it will be a lot more expensive (with that i860 on board). Intel is not the only company with RISC technology, Motorola could just as easily slap an 88k onboard the 050. Still, I don't like this sort of speculation. Motorola may only be discussing the 050 now, I bet designs aren't even on the board yet, so they could still take any path from here. Intel and Motorola have always seem to leapfrog each other in the past and chip performance has been about equal (486 and 040 Specmarks are about the same, with a slight integer performance advantage on the 486 side, and a slight floating point advantage on the 040 side) What you may want to consider is, is this the end of CISC? I can't say. I remember a few years ago when the same question was asked, and then the CISC camp came back fighting with better chips (internal pipelining, onboard cache, and faster instruction timings). It seems RISC is back in the lead again. > My guess is that Motorola has been told by its major 68030/040 users >that they're ready to switch away from the 680x0 family [e.g. NeXT, >Apple with possibly the 88K, HP with their own PA RISC] for their >high-end products, and so Motorola is not putting too much effort into >designing high-end follow-ups to the 68040. > > What do you think? Will the 68050 be the last high-end member of >the 680x0 line? I really don't think so. Intel and Motorola's bread and butter in the chips department come from the sales of vast quantities of the low-end stuff. I think it's dangerous to change processors on machines like PC's, Mac's, Amiga's, etc. You usually don't get source code with software on these machines, so changing the processor will probably fail when 100% of all previous software won't work on it. It's the same effect that would happen if Apple did something like discard the MacOS in favor of a new one (like AmigaDOS :) , yeah it's a joke) A P-CODE executable loader system makes softare CPU independent unfortunately neither Amiga, Mac, or any of the other personal computers support a system like this. Let's face it, the major selling point of the Mac is it's OS and software. If Apple enters the high-end workstation market with a new processor and architecture, they will be starting basically scratch. I wouldn't want CBM to make the same mistake either. (I don't think the A3000UX fits into thhis category. I think it fits into the A/UX category since the A3000UX can run Amiga software, and is still basically an Amiga.) Besides, I think Apple might ruin the concept of a workstation by making it Macish. Workstations need to be preemptive multitasking, multiuser, memory protected, resource tracking, virtual memory, lots of disk space, and usuable from tty's and shells. If Apple wants to spend their cash, let them spend it on better manufacturing facilities (totally automated like the NeXT, and all the factories in Japan) so prices are cheaper. They could also spend time working on multiprocessing and perhaps multiprocessing over Macintosh networks. Be innovative and add to the computer community rather than trying to enter a market that is already populated with nice workstations trying to suck up more cash. I don't think Apple would do anything in the workstation market except take existing technology and chips, slap them into a new machine and try to sell it with a large advertising campaign as an innovative workstation.(They would probably copyright it too, and sue other workstation vendors :) ) (oh yes, I know the 88k has loadable microcode) >-- >------------------------------------------------------------------------------ >Evan Torrie. Stanford University, Class of 199? torrie@cs.stanford.edu >"If it weren't for your gumboots, where would you be? You'd be in the >hospital, or in-firm-ary..." F. Dagg -- / INET:rjc@gnu.ai.mit.edu * // The opinions expressed here do not \ | INET:r_cromwe@upr2.clu.net | \X/ in any way reflect the views of my self.| \ UUCP:uunet!tnc!m0023 * /
peterk@cbmger.UUCP (Peter Kittel GERMANY) (06/10/91)
In article <1991Jun10.072945.8821@neon.Stanford.EDU> torrie@cs.stanford.edu (Evan Torrie) writes: > > What do you think? Will the 68050 be the last high-end member of >the 680x0 line? No, I think the 68060 will be the end of the line. Why? Because there already is a 68070 existing (a 68010 plus some peripheral built in, made by Philips in license) :-) -- Best regards, Dr. Peter Kittel // E-Mail to \\ Only my personal opinions... Commodore Frankfurt, Germany \X/ {uunet|pyramid|rutgers}!cbmvax!cbmger!peterk
es1@cunixb.cc.columbia.edu (Ethan Solomita) (06/10/91)
In article <1991Jun10.072945.8821@neon.Stanford.EDU> torrie@cs.stanford.edu (Evan Torrie) writes: > > What do you think? Will the 68050 be the last high-end member of >the 680x0 line? > Sounds to me like they have found a number of (comparatively) easy changes they can make to the 040 to speed it up and they want to do them first, before they make major changes. That announcement doesn't guarantee things either way. -- Ethan Now the world has gone to bed, Now I lay me down to sleep, Darkness won't engulf my head, Try to count electric sheep, I can see by infrared, Sweet dream wishes you can keep, How I hate the night. How I hate the night. -- Marvin
chucks@pnet51.orb.mn.org (Erik Funkenbusch) (06/11/91)
torrie@cs.stanford.edu (Evan Torrie) writes: > Actually, reading between the lines, this announcement sounds like >the beginning of the end for the 680x0 line. The article says that >the '050 will be an evolutionary upgrade, using just improved process >fabrication and a few small architectural changes to get its >increased speed. All in all, it sounds very much like the situation >with the 68030 vs 68020 - claimed "double the performance" with an >evolutionary upgrade, which actually ended up with only 25-30% >improvement. Perhaps an 030 on a MAC is 25% speedup. but i had a 020 machine before this one and i only recieved about 3% the speed of a stock Amiga. when i got the 030 in this one, i shot up to ever 10x the speed of a stock Amiga.. that's hardly a 25% increase. the ONLY difference was that the 020 was 16Mhz and the 030 is 25Mhz. but that only accounts for approximately 25% increase. just speeding up the processor. > > Why do I see this as the beginning of the end? Well, by all accounts, >Intel's 80586 will be as revolutionary a chip from the 486 as the 386 >was from the 286. Reports of an i860 on board, superscalar processing >etc seem to dominate any discussion of the 586. > > If Motorola is just planning an evolutionary upgrade for the 050, I >can't see how this is going to compete head-to-head with the 586. All this is just talk.. as none of us KNOW what each new chip will be yet. i hardly doubt that even Intel OR Motorola will know untill they finalize the design. > > My guess is that Motorola has been told by its major 68030/040 users >that they're ready to switch away from the 680x0 family [e.g. NeXT, >Apple with possibly the 88K, HP with their own PA RISC] for their >high-end products, and so Motorola is not putting too much effort into >designing high-end follow-ups to the 68040. > > What do you think? Will the 68050 be the last high-end member of >the 680x0 line? I hardly think that any of Motorola's MAJOR players are telling them they will swith to 88000's... Since i would imagine that Amiga/Mac/Atari/NeXT/etc.. are the bread and butter of the 68XXX line.. with the Unix stations pulling up after most of those.. just how many 68XXX base Unix Stations are being sold these days? not many. .--------------------------------------------------------------------------. | UUCP: {amdahl!tcnet, crash}!orbit!pnet51!chucks | "I know he's come back | | ARPA: crash!orbit!pnet51!chucks@nosc.mil | from the dead, but do | | INET: chucks@pnet51.orb.mn.org | you really think he's | |-------------------------------------------------| moved back in?" | | Amiga programmer at large, employment options | Lou Diamond Philips in | | welcome, inquire within. | "The First Power". | `--------------------------------------------------------------------------'
cg@ami-cg.UUCP (Chris Gray) (06/11/91)
In article <1991Jun10.083905.9329@mintaka.lcs.mit.edu> rjc@wookumz.gnu.ai.mit.e > Still, I don't like this sort of speculation. Motorola may only >be discussing the 050 now, I bet designs aren't even on the board >yet, They're quite a bit beyond that stage, from rumors I've heard. It takes a LOT of work to bring such a thing to a sellable state. I would image that initial overall design stuff for a "68050" was done at least 2 years ago. I haven't the faintest idea what's in the thing, however. It would be nice if they could cram the rest of the FPU back in! -- Chris Gray alberta!ami-cg!cg or cg%ami-cg@scapa.cs.UAlberta.CA
torrie@cs.stanford.edu (Evan Torrie) (06/11/91)
rjc@wookumz.gnu.ai.mit.edu (Ray Cromwell) writes: >In article <1991Jun10.072945.8821@neon.Stanford.EDU> torrie@cs.stanford.edu (Evan Torrie) writes: >> >> Actually, reading between the lines, this announcement sounds like >>the beginning of the end for the 680x0 line. The article says that >>can't see how this is going to compete head-to-head with the 586. > Let's not engage in speculation of what the 050 will or will not be >based on a blurb. Hey, but this is .advocacy! We're supposed to make wild speculation here about unannounced products, aren't we? >the 050 if it is to come out in only a year from now, will probably be >an evolutionary upgrade like you said, The information in the article is quoted from Jim Reinhart, Motorola manager of marketing and applications for the 680x0 family. "Reinhart called the 68050 an 'evolutionary' product, in that much of its technology is derived from the '040. But some modifications in architecture and improvements in manufacturing will double performance at a given clock speed, he said. The response from customers has been 'very enthusiastic,' he said. >however, if Intel is planning >such drastic architecture changes the 586 may take a lot longer >and it will be a lot more expensive (with that i860 on board). >Intel is not the only company with RISC technology, Motorola >could just as easily slap an 88k onboard the 050. Yes, except I think that unlikely. Consider: Intel has been quoted in the press (specifically EE Times) as confirming that the 80x86 will be their processor of choice for the rest of the 90's. Have you heard Motorola say the same about the 680x0? In fact, if you look at what the 88K people at Motorola have been saying, it looks like they plan the 88K to be their processor of the 90's. > Still, I don't like this sort of speculation. Motorola may only >be discussing the 050 now, I bet designs aren't even on the board >yet, so they could still take any path from here. If Motorola plans to have the 050 out next year, I would hope that they have had real designs for at least the past year! For complicated CISC chips like these, it seems to take at least 3 years development time. >> What do you think? Will the 68050 be the last high-end member of >>the 680x0 line? > I really don't think so. Intel and Motorola's bread and butter in the >chips department come from the sales of vast quantities of the low-end >stuff. I agree, and I expect Motorola to continue to sell a lot of 68030's, and 68040's through the rest of this decade. But do you think that they will bother to do any more HIGH-END members beyond the 050? I have my doubts. >I think it's dangerous to change processors on machines >like PC's, Mac's, Amiga's, etc. You usually don't get source code >with software on these machines, so changing the processor will probably >fail when 100% of all previous software won't work on it. Unless you add in some sort of compatibility box in the interim while new software is being developed. >It's the same >effect that would happen if Apple did something like discard the MacOS >in favor of a new one (like AmigaDOS :) , yeah it's a joke) Well, they're almost certainly going to do this with next years's (??) post-Mac generation. [but they'll retain a Mac software compatibility until new software is developed]. >Mac is it's OS and software. If Apple enters the high-end workstation >market with a new processor and architecture, they will be starting basically >scratch. See above. >Besides, I think Apple might ruin the concept of >a workstation by making it Macish. Workstations need to be >preemptive multitasking, multiuser, memory protected, resource tracking, >virtual memory, lots of disk space, and usuable from tty's and shells. Hey! Sounds just like A/UX! >populated with nice workstations trying to suck up more cash. I don't >think Apple would do anything in the workstation market except >take existing technology and chips, slap them into a new machine >and try to sell it with a large advertising campaign as an innovative >workstation. I hope Apple doesn't do this, because I don't think they'll sell a lot of machines that way, competing as a Unix box seller. And from the rumours, it sounds as though they probably won't be using Unix anyway. I hope that they'd instead aim their RISC machines as just high-end personal computing machines, using the increased power to provide an even nicer user interface [e.g. voice recognition, 3-D graphics, object-oriented operating system like PenPoint etc.] -- ------------------------------------------------------------------------------ Evan Torrie. Stanford University, Class of 199? torrie@cs.stanford.edu "If it weren't for your gumboots, where would you be? You'd be in the hospital, or in-firm-ary..." F. Dagg
torrie@cs.stanford.edu (Evan Torrie) (06/11/91)
chucks@pnet51.orb.mn.org (Erik Funkenbusch) writes: >I hardly think that any of Motorola's MAJOR players are telling them they will >swith to 88000's... Since i would imagine that Amiga/Mac/Atari/NeXT/etc.. are ^^^^ Take NeXT out of here and put it in the Unix workstation group. >the bread and butter of the 68XXX line.. Yes, these machines are the bread and butter, and I think they'll continue to sell a LOT of low-end 680x0 CPUs for this market [e.g. I can see a 68050 Mac lasting until 1996 without too many problems]. But, I'm talking about the high-end. What incentive is there for Motorola to produce a 68060 if all their high-end users [such as NeXT, HP, and you could say Apple] have switched to another processor family? Maybe Motorola will do it for Commodore and Atari if they haven't switched by then [or at least convinced everyone that their 80586 machines are really far better than the Amiga or ST - yuck, what a horrible thought]. >with the Unix stations pulling up >after most of those.. just how many 68XXX base Unix Stations are being sold >these days? not many. Not many, but these are the machines which are using up all the high-end 68040's. -- ------------------------------------------------------------------------------ Evan Torrie. Stanford University, Class of 199? torrie@cs.stanford.edu "If it weren't for your gumboots, where would you be? You'd be in the hospital, or in-firm-ary..." F. Dagg
tegen@isy.liu.se (Claes Tegenf{lt) (06/11/91)
torrie@cs.stanford.edu (Evan Torrie) writes: >>blitter to your computer). The 50MHz 040 is due in 1Q 92. The >>article also says that the 68050 will be available sometime next year. >>It will double the performance of the 040 at the same clock speed. .. > If Motorola is just planning an evolutionary upgrade for the 050, I >can't see how this is going to compete head-to-head with the 586. .. > My guess is that Motorola has been told by its major 68030/040 users >that they're ready to switch away from the 680x0 family [e.g. NeXT, >Apple with possibly the 88K, HP with their own PA RISC] for their >high-end products, and so Motorola is not putting too much effort into >designing high-end follow-ups to the 68040. If you have a company and some of your old customers threatened you with starting to buy someone else's products, would you really just sit down and wait until they do? That is what you propose Motorola will do... I think, IF customers threaten Motorola by suggesting their product line 680x0 isn't enough, THEN they will double their efforts to develop 68050, 68060, etc. I think the 680x0 series has every chance in the world not just to survive, but to gain in popularity! And I also would like to propose that the 586 processor shouldn't be compared with the 68050, but rather to 68040, then they may even compete! ;-> Clas Tegenfeldt Linkoping University tegen@isy.liu.se -- ------------------------------------------------------------------------- | Clas Tegenfeldt | tegen@isy.liu.se | "Magic cooking software". | -------------------------------------------------------------------------
chucks@pnet51.orb.mn.org (Erik Funkenbusch) (06/12/91)
I recall reading in the VERY FIRST major announcement of the 040 (Byte, january '90) that said they were already working on the 050. It seems to me that Moto might have originally intened the 040 to be what the 050 will be. but technology restraints may have forced them to limit the 040 until the proper technology was developed to do what they wanted. for instance, you can only make circuit paths so small today and be reliable. .--------------------------------------------------------------------------. | UUCP: {amdahl!tcnet, crash}!orbit!pnet51!chucks | "I know he's come back | | ARPA: crash!orbit!pnet51!chucks@nosc.mil | from the dead, but do | | INET: chucks@pnet51.orb.mn.org | you really think he's | |-------------------------------------------------| moved back in?" | | Amiga programmer at large, employment options | Lou Diamond Philips in | | welcome, inquire within. | "The First Power". | `--------------------------------------------------------------------------'
daveh@cbmvax.commodore.com (Dave Haynie) (06/13/91)
In article <1991Jun10.072945.8821@neon.Stanford.EDU> torrie@cs.stanford.edu (Evan Torrie) writes: >melling@cs.psu.edu (Michael D Mellinger) writes: >All in all, it sounds very much like the situation with the 68030 vs 68020 >- claimed "double the performance" with an evolutionary upgrade, which >actually ended up with only 25-30% improvement. If you're building full speed systems, a 68030 is _better_ than twice as fast at the same clock speed as the equivalent 68020/68851 combination. If you simply drop a 68030 into a 68020 system, the increase is more like 15%-20% (the Mac II vs. IIx|IIcx is the only pure example of this I know of; the IIci and IIfx, A2630, and all the HP 68030 systems are designed as 68030 systems). > Why do I see this as the beginning of the end? Well, by all accounts, >Intel's 80586 will be as revolutionary a chip from the 486 as the 386 >was from the 286. Who's accounts. Since the leap from 68030 to 68040 was greater than the leap from 80386 to 80486, it sure looks like Motorola's in the lead. I guess you can really pick the rumors you like until real information is released. If you keep in mind all the rumors that were around prior to the '486 and '040 introductions (the '486 "downloadable microcode, emulates any CPU at full speed" was my favorite "obviously wigged out" one). >Reports of an i860 on board, superscalar processing etc seem to dominate any >discussion of the 586. For two good reasons: [a] if you're not satisfied with a '486, you aren't likely going to be satisfied with a '586, so you like to hear about real speedups on the way, not just 2x improvement, and [b] Intel has been pushing the idea of using an i860 as a coprocessor, depite the fact it's not all that well suited to the job (thought the new one they just announced is better), simply because Intel would really, really, really, really like to get two of their expensive chips into every high end PClone, rather than just one. > If Motorola is just planning an evolutionary upgrade for the 050, I >can't see how this is going to compete head-to-head with the 586. Since you really don't know what either chip is going to do, I suggest it's far too early to start worrying about things. Also, keep in mind that Intel has much less to lose by wimping out on the '586 than Motorola does on the '050. No matter what they do, Intel will sell as many '586, '586SX (the one with the FPU disabled), '587SX, '588RC (Really Crippled), etc. as they can churn out. Motorola needs to worry about their market jumping ship to RISC processors, since the "needs to be binary-compatible" personal systems comprise much less of the Motorola market. And the 680x0 environment is a heck of a lot easier to emulate in software than the brain dead combination of MS-DOS and a '486 type CPU. > My guess is that Motorola has been told by its major 68030/040 users >that they're ready to switch away from the 680x0 family [e.g. NeXT, >Apple with possibly the 88K, HP with their own PA RISC] for their >high-end products, and so Motorola is not putting too much effort into >designing high-end follow-ups to the 68040. Considering the number of processors Motorola sells to non-desktop computer users, like VME and other industrial applications, the Apple/NeXT/Amiga use of high end 680x0s, while not trivial, is the minority. Incidently, the Enquirer, er, EE Times, rumors that Apple may be looking at IBM's "America" or whatever they call it, the processor set in the RS/6000, rather than 88K. IBM gets to port MacOS, they get a fast processor. Not that anyone, other than Apple and IBM, have a real clue at this point. -- Dave Haynie Commodore-Amiga (Amiga 3000) "The Crew That Never Rests" {uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy "This is my mistake. Let me make it good." -R.E.M.
daveh@cbmvax.commodore.com (Dave Haynie) (06/13/91)
In article <1338@cbmger.UUCP> peterk@cbmger.UUCP (Peter Kittel GERMANY) writes: >In article <1991Jun10.072945.8821@neon.Stanford.EDU> torrie@cs.stanford.edu (Evan Torrie) writes: >> What do you think? Will the 68050 be the last high-end member of >>the 680x0 line? >No, I think the 68060 will be the end of the line. Why? Because there >already is a 68070 existing (a 68010 plus some peripheral built in, >made by Philips in license) :-) Don't count on it. Signetics calling their part "68070" isn't far from some Amiga 3rd party calling their A500 add-on gizmo "A4000". Motorola and Commodore, respectively, have no particular need to avoid those names if they find them useful. -- Dave Haynie Commodore-Amiga (Amiga 3000) "The Crew That Never Rests" {uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy "This is my mistake. Let me make it good." -R.E.M.
torrie@cs.stanford.edu (Evan Torrie) (06/13/91)
daveh@cbmvax.commodore.com (Dave Haynie) writes: >In article <1991Jun10.072945.8821@neon.Stanford.EDU> torrie@cs.stanford.edu (Evan Torrie) writes: >If you're building full speed systems, a 68030 is _better_ than twice as fast >at the same clock speed as the equivalent 68020/68851 combination. If you >simply drop a 68030 into a 68020 system, the increase is more like 15%-20% >(the Mac II vs. IIx|IIcx is the only pure example of this I know of; the >IIci and IIfx, A2630, and all the HP 68030 systems are designed as 68030 >systems). I've never seen twice the performance in CPU intensive applications. Just as an example, the Speedometer benchmark on an original Mac II clocks the CPU as 3.49, and my IIci as 6.71. Adjusting for clock speed, that's only a 25% improvement. >> Why do I see this as the beginning of the end? Well, by all accounts, >>Intel's 80586 will be as revolutionary a chip from the 486 as the 386 >>was from the 286. >Who's accounts. EE Times, InfoWorld, PC Week for example. [Not exactly paragons of journalistic integrity, but they're more often right than wrong]. >>Reports of an i860 on board, superscalar processing etc seem to dominate any >>discussion of the 586. Well, also because of the reports that Intel is consolidating behind the 80x86 family, and trying to bundle their two other CPU development teams [the 860 (high-speed floating-point), and the 960 (superscalar RISC)] in behind the 80x86 team. >> If Motorola is just planning an evolutionary upgrade for the 050, I >>can't see how this is going to compete head-to-head with the 586. >Since you really don't know what either chip is going to do, I suggest it's >far too early to start worrying about things. We have some hints from Motorola (i.e. the "evolutionary" upgrade). >Also, keep in mind that Intel >has much less to lose by wimping out on the '586 than Motorola does on the >'050. This depends on whether you consider the ACE consortium to be a serious force. In fact, if you recall, one of the main thrusts of ACE is to allow Compaq to force Intel to increase the speed of the 80x86 family, or else risk losing the high-end market to the R4000. >Considering the number of processors Motorola sells to non-desktop computer >users, like VME and other industrial applications, the Apple/NeXT/Amiga use >of high end 680x0s, while not trivial, is the minority. Even here though, the 88K is getting the push ahead of the 68K. Take yesterday's Ford/Motorola announcement, for example. -- ------------------------------------------------------------------------------ Evan Torrie. Stanford University, Class of 199? torrie@cs.stanford.edu "Apes evolved from creationists" - seen on a bumper sticker.
daveh@cbmvax.commodore.com (Dave Haynie) (06/13/91)
In article <1991Jun13.003707.19785@neon.Stanford.EDU> torrie@cs.stanford.edu (Evan Torrie) writes: >daveh@cbmvax.commodore.com (Dave Haynie) writes: >>In article <1991Jun10.072945.8821@neon.Stanford.EDU> torrie@cs.stanford.edu (Evan Torrie) writes: >>If you're building full speed systems, a 68030 is _better_ than twice as fast >>at the same clock speed as the equivalent 68020/68851 combination. > I've never seen twice the performance in CPU intensive applications. >Just as an example, the Speedometer benchmark on an original Mac II clocks >the CPU as 3.49, and my IIci as 6.71. Adjusting for clock speed, that's only >a 25% improvement. I don't have a clue about what Speedometer does to get its figures. I can pick Amiga benchmarks which claim the A3000 is anywhere from 2x to 10x the speed of an A2500/20. It all depends on what you're testing. Also, while the A3000 or Mac IIci are decent 68030 systems, consider that on both machines, the DRAM is still running less that 1/2 the maximum speed that the '030 is capable of. Burst mode helps some, but it's no substitute for 0 wait state memory. Cache is what you want, you aren't going to get DRAM going quite that fast (well, maybe a two-way interleaved 60ns memory system could give you 1 wait state...). Keep in mind that memory speed doesn't scale with CPU speed; the 68020/68851 system didn't use that much slower memory. All of this point out that it's difficult to build a full speed 68030 system. Basically, you don't for the budget me and Greg had for the A3000 and I imagine the Apple folks had for the IIci. HP did some killer, full speed 68030 systems, but they cost relatively big bucks. It's easier to build a 68040 system that goes close to full speed, since [a] the cache hits considerably more often, so external memory speed isn't quite as critical, and [b] the 68040 bus interface is just plain better -- you can run a whole clock faster with 80ns DRAM at 25MHz with the '040. >>> Why do I see this as the beginning of the end? Well, by all accounts, >>>Intel's 80586 will be as revolutionary a chip from the 486 as the 386 >>>was from the 286. >>Who's accounts. > EE Times, InfoWorld, PC Week for example. [Not exactly paragons of >journalistic integrity, but they're more often right than wrong]. I read EE Times on the can every week and didn't get that impression. They did point out that Intel is throwing most of their efforts behind 80x86s, RISCy things are now officially just a sideline. I guess Intel knows where their money is grown. The 80x86 people are a captive market, so Intel can gouge pretty heavily. Other companies make RISC chips. > Well, also because of the reports that Intel is consolidating behind the >80x86 family, and trying to bundle their two other CPU development teams >[the 860 (high-speed floating-point), and the 960 (superscalar RISC)] >in behind the 80x86 team. They're continuing to push the i860 and i960 into their established niches. The i860 as a graphics coprocessor and vector machine (it is showing up on all kinds of high end display boards), the i960 for embedded control, especially in the military and avionics markets. The i960 has nothing at all to do with the 80x86 line. They will have some kind of glue chip to make it easier to hook an i860 up to an 80x86 system. There's that second big bucks chip in the same box, as I mentioned before. > We have some hints from Motorola (i.e. the "evolutionary" upgrade). Where's the line between "evolutionary" and "revolutionary"? The 68020->68030 upgrade is generally considered evolutionary, although the speedup was 2x for fully outfitted, equivalent systems. Most people consider the i386->i486 move to be something closer to revolutionary, even though that's basically just around 2x improvement. Architecturally speaking, the i486 upgrade was much more of a revolution, but then again, the '040 was an even larger break from the '030. Anyway, looking at the Motorola track record, the '050 should be in the "evolutionary" category, which is tracked more on architectural grounds than performance upgrades: 68000,68010 16 bit 68020,68030 32 bit 68040,68050 Cool 32 bit Except for the 68000 to 68010 upgrade, each step so far has been at least a factor of two improvement. >>Also, keep in mind that Intel has much less to lose by wimping out on the >>'586 than Motorola does on the '050. > This depends on whether you consider the ACE consortium to be a serious >force. Well, no one else does :-) (EE Times was lukewarm on it, Microprocessor Report pretty unimpressed, since it doesn't really mean much of anything yet). >In fact, if you recall, one of the main thrusts of ACE is to allow Compaq to >force Intel to increase the speed of the 80x86 family, or else risk losing >the high-end market to the R4000. No, actually, it's Compaq trying to get out from behind Intel and IBM with something they can control. And it makes sense. They could fall in line behind some of the SparcStation cloners, but that puts them in a non-controlling position, just like they are in the Intel market. Actually worse, since Sun, IEEE support or not, has controlled the SPARC architecture, basic system design, and software. At least in the Intel market, you had that split three ways: Intel on the chip architecture, IBM on the system design, and Microsoft on the software. Worked good. And guess what? ACE gives you the same thing: MIPS on the chip architecture, Compaq|DEC on the system architecture, and Microsoft on the software. So, basically, I think ACE is an interesting but pretty enivitable happening (though relatively little has actually happened yet). No matter what Intel does, Compaq is likely to build ARC systems, assuming the ACE consortium can actually pull this whole thing off. >>Considering the number of processors Motorola sells to non-desktop computer >>users, like VME and other industrial applications, the Apple/NeXT/Amiga use >>of high end 680x0s, while not trivial, is the minority. > Even here though, the 88K is getting the push ahead of the 68K. Take >yesterday's Ford/Motorola announcement, for example. Sure, the 68K isn't for everything in that market. Not only that, but it's been so successful in that market that Motorola needs to make sure the 88K follows in its footsteps. But really, the 68K is alive and kicking, especially in VME. The 88K is on VME, but hasn't really gone far yet. Take a look through Computer Design sometime and check out all the '040 based VME cards out already. Ultimately, both Intel and Motorola have to "put up or shut up" with their CISC products. At present, anyway, it's easier for more Motorola customers to abandon the 68K than for Intel users to abandon the 80x86. Not that either is "easy" by any stretch of the imagination. And there's no guarantee that, when anyone abandons a 68K/80x86, that either Motorola or Intel will get the business. In fact, Intel's just about assuring that, if you give up the 80x86, you won't replace it with an Intel chip, as they're pushing their RISCs into niche markets. Motorola, I think wisely, is building the 88K to do the same things the 68K does. However, there's absolutely no assurance that a 68K user will necessarily pick up the 88K as a replacement; MIPSs and SPARCs, to name a few, are out now for both system CPUs, VME stuff, and embedded control. -- Dave Haynie Commodore-Amiga (Amiga 3000) "The Crew That Never Rests" {uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy "This is my mistake. Let me make it good." -R.E.M.
rkushner@sycom.UUCP (Ronald Kushner) (06/14/91)
torrie@cs.stanford.edu (Evan Torrie) writes: > EE Times, InfoWorld, PC Week for example. [Not exactly paragons of >journalistic integrity, but they're more often right than wrong]. Do these magazines really rag on the 486sx/487sx combo, or are they calling it a triumph to the little guy, who can finally afford an 486, and then can fork over a few dollars for the mathco if they like? It would give you some insight on their bias... -- C-UseNet V0.42d Ronald Kushner Life in Hell BBS +1 (313) 939-6666 P.O. Box 353 14400 USR HST V.42 & V.42bis Sterling Heights, MI 48311-0353 Complete Amiga Support UUCP: uunet!umich!vela!sycom!rkushner (We are not satanic, just NUTS!) DISCLAIMER: I say what I mean, and mean what I say.
torrie@cs.stanford.edu (Evan Torrie) (06/14/91)
daveh@cbmvax.commodore.com (Dave Haynie) writes: >Burst mode helps some, but it's no substitute for 0 wait state memory. Cache Wouldn't this also be true for a 68020 system? I'd be interested to see benchmarked differences between HP 68020 and 68030 systems. >> Well, also because of the reports that Intel is consolidating behind the >>80x86 family, and trying to bundle their two other CPU development teams >>[the 860 (high-speed floating-point), and the 960 (superscalar RISC)] >>in behind the 80x86 team. >They're continuing to push the i860 and i960 into their established niches. >The i860 as a graphics coprocessor and vector machine (it is showing up on >all kinds of high end display boards), the i960 for embedded control, >especially in the military and avionics markets. Yes, but I meant that Intel would use the "expertise" gathered from developing these processors in designing the 586. Up till now, the teams working on these processors have been kept pretty much separate. The articles I've been reading indicate that these people are now working much more closely. >>>Also, keep in mind that Intel has much less to lose by wimping out on the >>>'586 than Motorola does on the '050. >> This depends on whether you consider the ACE consortium to be a serious >>force. >Well, no one else does :-) (EE Times was lukewarm on it, Microprocessor Report >pretty unimpressed, since it doesn't really mean much of anything yet). "Yet" is the important word. Sure, it's all vapourware at the moment, but I think the point of announcing it now was to scare Intel. If R4000 machines really come to market in a big way, I'm sure Intel will try be trying very hard to make their 80x86 family competitive to win back that high end [which in a few years time will be the low-end]. >>In fact, if you recall, one of the main thrusts of ACE is to allow Compaq to >>force Intel to increase the speed of the 80x86 family, or else risk losing >>the high-end market to the R4000. >No, actually, it's Compaq trying to get out from behind Intel and IBM with >something they can control. I don't see why Compaq will be able to control this any better than the 80x86 market, if they're agreeing to all these other firms as ACE members. DEC, for example, could just as easily tell the market where to go [especially as DEC has much more experience in the high end than Compaq]. And who says DEC will do what Compaq wants? [DEC, for example, could make TurboChannel the preferred bus, leaving EISA out in the cold]. I still think Compaq would prefer to stay with the 80x86 family [at least that's the impression I got as to why ACE included the 80x86 as a valid architecture, rather than sticking strictly with the R4000]. >follows in its footsteps. But really, the 68K is alive and kicking, >especially in VME. The 88K is on VME, but hasn't really gone far yet. Take a >look through Computer Design sometime and check out all the '040 based VME >cards out already. I know, I read it quite often. But that doesn't say which one Motorola is actually putting their effort behing. I'd be interested to know the relative size of the 68K and 88K R&D groups, along with their growth over the past 4 years. >do the same things the 68K does. However, there's absolutely no assurance >that a 68K user will necessarily pick up the 88K as a replacement; MIPSs and >SPARCs, to name a few, are out now for both system CPUs, VME stuff, and >embedded control. Current 68K users are unlikely to go the SPARC route, given its poor performance and relatively poor architecture. [You'll note that a lot of 80x86 vendors are now selling SPARC systems - I guess they know how to pick a bad one :-)]. MIPS is definitely Motorola's biggest threat, especially if Motorola doesn't get the 88110 out and shipping soon after the R4000. I think Motorola will make it extremely attractive for current 68K users to switch to switch to the 88K (i.e. in financial measures) - this seems to be the approach taken with NeXT. -- ------------------------------------------------------------------------------ Evan Torrie. Stanford University, Class of 199? torrie@cs.stanford.edu Fame, fame, fame... What's it good for? Ab-so-lute-ly nothing
daveh@cbmvax.commodore.com (Dave Haynie) (06/15/91)
In article <1991Jun14.004412.26009@neon.Stanford.EDU> torrie@cs.stanford.edu (Evan Torrie) writes: >daveh@cbmvax.commodore.com (Dave Haynie) writes: >>Burst mode helps some, but it's no substitute for 0 wait state memory. Cache > Wouldn't this also be true for a 68020 system? No. The 68020 doesn't support burst mode. It also has a minimal cycle time of three clocks. If you add the 68851 MMU (like A2620 or original Mac II, for 68030 equivalence), the minimum cycle time becomes four clocks. The 68030's minimum memory cycle time is 2 clocks (burst averages 1.25 clocks, but it can do unnecessary fetches). >>They're continuing to push the i860 and i960 into their established niches. >>The i860 as a graphics coprocessor and vector machine (it is showing up on >>all kinds of high end display boards), the i960 for embedded control, >>especially in the military and avionics markets. > Yes, but I meant that Intel would use the "expertise" gathered from >developing these processors in designing the 586. That's no different that Motorola using what it learns on the 96002 and 88000 for future 680x0 family members. In fact, there's more than a passing resembelence between the 68040 and the 88100/88200 in a number of areas. Rumor has it the on-chip 88110 cache will be as fast as the 68040's (currently 88200s hit in two clocks, like the 68030, rather than the single clock of the 68040). >>Well, no one else does :-) (EE Times was lukewarm on it, Microprocessor Report >>pretty unimpressed, since it doesn't really mean much of anything yet). > "Yet" is the important word. Sure, it's all vapourware at the moment, >but I think the point of announcing it now was to scare Intel. I really don't think so. No one in the group has a good reason to scare Intel. You might argue for Compaq wanting better Intel CPUs, maybe they're not so concerned if they're committed to ACE. DEC and SGI certainly have no motive in getting Intel excited, neither do Microsoft or MIPS. >>No, actually, it's Compaq trying to get out from behind Intel and IBM with >>something they can control. > I don't see why Compaq will be able to control this any better than the >80x86 market, if they're agreeing to all these other firms as ACE members. Compaq is one of the leaders. They're getting their two cents in now, at the start, and making it happen. Compaq and DEC are in the IBM position themselves with this. Even if it's set up as a cooperative rather than "IBM and Intel do, the rest of you follow", Compaq's better off. Also, keep in mind that Compaq is selling at the high end of the PClone market now. They are having problems with cheap, lower end systems just like IBM. It's real obvious they're not going to turn around and try to be a Tandy or Tawian, Inc. So moving upward is a logical thing for them. Especially a move to the R4000, which Intel can't possibly catch with any new 80x86 processor (heck, the best '486s don't get near today's R3001s). > I still think Compaq would prefer to stay with the 80x86 family >[at least that's the impression I got as to why ACE included the 80x86 >as a valid architecture, rather than sticking strictly with the R4000]. I think that was done, more than anything, for software reasons. If you make high end PClones ACE compatible, at the source level (actually, it's more like specifying the ACE system as being compatible in the first place, since this Intel stuff already exists), then you get a free flow of software from the extremely large PClone market, since porting to the MIPS machines will be a piece of cake. > Current 68K users are unlikely to go the SPARC route, given its poor >performance and relatively poor architecture. [You'll note that a lot >of 80x86 vendors are now selling SPARC systems - I guess they know how >to pick a bad one :-)]. I guess Sun would be a little surprised to hear that, since they replaced the 68K with SPARC, and in fact, really hadn't quite "taken off" as a major force in the workstation market until they did. But I do agree that the SPARC architecture is weak when compared to the current alternatives. Clone makers don't care about that, they're not going to go out and build a brand new computer based on 88110 or R4000 or any merchant PA-RISC that they can eventually do. They're going to copy something. SPARC machines are faster than 80x86s, and Sun's making it easy for these guys to make these copies (of course, not real fast ones, but the low end systems). >MIPS is definitely Motorola's biggest threat, especially if Motorola doesn't >get the 88110 out and shipping soon after the R4000. I agree there, they are the two I would look at if I had to start building a RISC system this year. -- Dave Haynie Commodore-Amiga (Amiga 3000) "The Crew That Never Rests" {uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy "This is my mistake. Let me make it good." -R.E.M.
rblewitt@sdcc6.ucsd.edu (Richard Blewitt) (06/15/91)
In article <rkushner.7503@sycom.UUCP> rkushner@sycom.UUCP (Ronald Kushner) writes: >torrie@cs.stanford.edu (Evan Torrie) writes: >> EE Times, InfoWorld, PC Week for example. [Not exactly paragons of >>journalistic integrity, but they're more often right than wrong]. > > >Do these magazines really rag on the 486sx/487sx combo, or are they calling >it a triumph to the little guy, who can finally afford an 486, and then can >fork over a few dollars for the mathco if they like? > >It would give you some insight on their bias... > Well, look at the latest BYTE, they really pan the 486sx, calling it not inovative at all and saying a few other things along that line. Rick _____________________________________________________________________ _____________________________.sig____________________________________ _____________________________________________________________________ The generic .sig Rick Blewitt rblewitt@ucsd.edu
chucks@pnet51.orb.mn.org (Erik Funkenbusch) (06/15/91)
torrie@cs.stanford.edu (Evan Torrie) writes: > I've never seen twice the performance in CPU intensive applications. >Just as an example, the Speedometer benchmark on an original Mac II clocks >the CPU as 3.49, and my IIci as 6.71. Adjusting for clock speed, that's only >a 25% improvement. Well, my benchmark programs for my old 2620 board showed it as 3.5x the speed of a stock Amiga, and my 2630 card (030) using the same benchmark get's 10x the speed of a stock Amiga. the 20 card is 16Mhz, the 30 is 25 Mhz, adjusting for speed that's about 2x the performance. Apple is NOTORIOUS for breaking things so that they can sell you a better version later. this is why i believe the first 040 macs will be 16 Mhz. so they can sell you a 25 Mhz later.. .--------------------------------------------------------------------------. | UUCP: {amdahl!tcnet, crash}!orbit!pnet51!chucks | "I know he's come back | | ARPA: crash!orbit!pnet51!chucks@nosc.mil | from the dead, but do | | INET: chucks@pnet51.orb.mn.org | you really think he's | |-------------------------------------------------| moved back in?" | | Amiga programmer at large, employment options | Lou Diamond Philips in | | welcome, inquire within. | "The First Power". | `--------------------------------------------------------------------------'
eachus@largo.mitre.org (Robert I. Eachus) (06/15/91)
In article <22393@cbmvax.commodore.com> daveh@cbmvax.commodore.com (Dave Haynie) writes: Anyway, looking at the Motorola track record, the '050 should be in the "evolutionary" category, which is tracked more on architectural grounds than performance upgrades: 68000,68010 16 bit 68020,68030 32 bit 68040,68050 Cool 32 bit I don't know what Cool 32 bit is, but I see the splits slightly differently. 68000, 68010: 24(27)-bit address, 16-bit data 68020: 32(36)-bit address, 32-bit data 68030, 68040: 32(36)-bit address, 32-bit data with cache fill support 68050: ???-bit address, 64(128-bit) data The 68050 is of course my WAG, but the easiest way to pump more performance out of the architecture without increasing the clock would be to add more data lines. (I'm sure the cache sizes will also increase, but the remaining advantage from that is in the noise.) Increasing the instruction and data path separation might help some, but if you are going to have more lines off the chip, might as well make them multipurpose by widening the the path instead of having two parallel paths. I think anyone who wants to significantly better the MIPS/clock of the current generation is going to have to go to wider busses like the RS/6000 and the HP Snake. I don't think that there will be any need for additional address lines in the 68050, but with the current generation having MMU's on board, allowing more physical memory than per user virtual memory is very cheap. There are machines out there NOW pushing the 4 Gig 32-bit addressing limit. It's just hasn't happened in the workstation market yet. Most of our workstations here are around 16 Meg, but we do have one Sun with 128 Meg, so the four Meg limit could become a problem in the 68050 time frame. (Although--if you know what you are doing--the current 680x0 scheme has 16 4 Meg address spaces, and a clever hardware design could support several user segments. But much better to do it right in the next chip and not get into Intel style kludges.) > Except for the 68000 to 68010 upgrade, each step so far has been at > least a factor of two improvement. Depends on what you are building and how you count! Because of the uninterruptable instruction problems in the 68000 (some instructions in the 68000 cannot be correctly restarted after a page fault), at Stratus it took twice as many 68000's as 68010's to make a logical CPU. (Eight 68000's or four 68010's if you care.) So for most people doing demand paged virtual memory, the 68010 was more than a factor of two improvement. :-) -- Robert I. Eachus with STANDARD_DISCLAIMER; use STANDARD_DISCLAIMER; function MESSAGE (TEXT: in CLEVER_IDEAS) return BETTER_IDEAS is...
torrie@cs.stanford.edu (Evan Torrie) (06/16/91)
daveh@cbmvax.commodore.com (Dave Haynie) writes: >In article <1991Jun14.004412.26009@neon.Stanford.EDU> torrie@cs.stanford.edu (Evan Torrie) writes: >You might argue for Compaq wanting better Intel CPUs, maybe they're not so >concerned if they're committed to ACE. DEC and SGI certainly have no motive >in getting Intel excited, neither do Microsoft or MIPS. Well, DEC does have a range of high-end 80x86 PC's (interestingly, just re-badged Intel machines). Although they're also much more committed to MIPS than Compaq, having already produced R3000 workstations. I still think Compaq and all the hanger-on clone makers in ACE have a good reason for wanting to put the wind up Intel. >> I still think Compaq would prefer to stay with the 80x86 family >I think that was done, more than anything, for software reasons. If you make >high end PClones ACE compatible, at the source level (actually, it's more >like specifying the ACE system as being compatible in the first place, since >this Intel stuff already exists), then you get a free flow of software from >the extremely large PClone market, since porting to the MIPS machines will >be a piece of cake. This has yet to be proved - i.e. it all depends on Microsoft actually getting a portable OS/2 v3.0, and including all the Windows stuff, and having a compiler which can compile to both architectures equally well. Microsoft's track record with advanced development hasn't been exactly impressive - check out their TrueImage debacle. >> Current 68K users are unlikely to go the SPARC route, given its poor >>performance and relatively poor architecture. [You'll note that a lot >>of 80x86 vendors are now selling SPARC systems - I guess they know how >>to pick a bad one :-)]. >I guess Sun would be a little surprised to hear that, since they replaced the >68K with SPARC, and in fact, really hadn't quite "taken off" as a major force >in the workstation market until they did. I think Sun went SPARC only because of their limited time-to-market. They wanted a RISC machine out as fast as possible, and the SPARC was a way to build a cheap, fast processor out of a gate array. Unfortunately, the architectural niceties were lost in the rush. >eventually do. They're going to copy something. SPARC machines are faster >than 80x86s, and Sun's making it easy for these guys to make these copies (of >course, not real fast ones, but the low end systems). Although they're not making it easy for them to sell them, [as in preventing their resellers from stocking non-Sun SPARC machines]. I have a feeling that Sun is shooting itself in the foot with SPARC. It's been over a year since these SPARC-clone machines were announced, and they've yet to make even a minimal dent. >>MIPS is definitely Motorola's biggest threat, especially if Motorola doesn't >>get the 88110 out and shipping soon after the R4000. >I agree there, they are the two I would look at if I had to start building a >RISC system this year. Or, as the rumours go, IBM's RS/6000 if it's available for licensing. :) -- ------------------------------------------------------------------------------ Evan Torrie. Stanford University, Class of 199? torrie@cs.stanford.edu "I didn't get where I am today without knowing a good deal when I see one, Reggie." "Yes, C.J."
torrie@cs.stanford.edu (Evan Torrie) (06/16/91)
chucks@pnet51.orb.mn.org (Erik Funkenbusch) writes: >Apple is NOTORIOUS for breaking things so that they can sell you a better >version later. this is why i believe the first 040 macs will be 16 Mhz. so >they can sell you a 25 Mhz later.. Can I quote you on this in about 4 months time? -- ------------------------------------------------------------------------------ Evan Torrie. Stanford University, Class of 199? torrie@cs.stanford.edu "I didn't get where I am today without knowing a good deal when I see one, Reggie." "Yes, C.J."
jerry@polygen.uucp (Jerry Shekhel) (06/16/91)
daveh@cbmvax.commodore.com (Dave Haynie) writes: > >Since the leap from 68030 to 68040 was greater than the leap >from 80386 to 80486, it sure looks like Motorola's in the lead >. What the hell are you talking about? The "leaps" in both product lines were exactly the same (on-board FPU, cache, speed improvements). Why do all of you Amigoids have it etched upon your brains that Motorola is, was, and always will be "clearly superior"? The difference in the technology has been slight ever since the '030/'386 (5 years now), especially if you're trying to look at the "big picture". Sure, Amiga and IBM clone technologies are pretty far apart, but their respective processors really aren't. > >Intel has been pushing >the idea of using an i860 as a coprocessor, depite the fact it's not all that >well suited to the job (thought the new one they just announced is better), >simply because Intel would really, really, really, really like to get two of >their expensive chips into every high end PClone, rather than just one. > Sorry Dave. The i860 is EXTREMELY well-suited to the job of a coprocessor, for graphics or floating point. It isn't that well-suited to the job of a CPU, however, because its integer unit is somewhat lacking, although a few manufacturers did it anyway. We'll have to see what happens with the new i860XL. You're right about Intel's marketing games, however. I cannot believe the ride they're offering customers now, with their 486SX/487SX processors. OK, I can accept the idea of the 486SX, for those people who don't need the FPU or cannot afford the full 486. But pricing the 487SX as a coprocessor (big bucks) is really sleazy, considering it is nothing more than a full 486 repackaged with an extra pin that turns off the 486SX. > >And the 680x0 environment is a >heck of a lot easier to emulate in software than the brain dead combination >of MS-DOS and a '486 type CPU. > MS-DOS is truly "braindead", but the '486 is not. In fact, it is an excellent 32-bit processor with a built in FPU, MMU, and a 64-terabyte virtual memory space. Show me how the '486 is braindead, Dave. Show me that you can do more, Dave, than post the same old ignorant Amigoid ramblings full of BS phrases like "braindead". > >Dave Haynie Commodore-Amiga (Amiga 3000) "The Crew That Never Rests" > {uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy > -- +-------------------+----------------------+---------------------------------+ | JERRY J. SHEKHEL | POLYGEN CORPORATION | When I was young, I had to walk | | Drummers do it... | Waltham, MA USA | to school and back every day -- | | ... In rhythm! | (617) 890-2175 | 20 miles, uphill both ways. | +-------------------+----------------------+---------------------------------+ | ...! [ princeton mit-eddie bu sunne ] !polygen!jerry | | jerry@polygen.com | +----------------------------------------------------------------------------+
mykes@amiga0.SF-Bay.ORG (Mike Schwartz) (06/16/91)
In article <1991Jun15.173528.3156@neon.Stanford.EDU> torrie@cs.stanford.edu (Evan Torrie) writes: >chucks@pnet51.orb.mn.org (Erik Funkenbusch) writes: > >>Apple is NOTORIOUS for breaking things so that they can sell you a better >>version later. this is why i believe the first 040 macs will be 16 Mhz. so >>they can sell you a 25 Mhz later.. > > Can I quote you on this in about 4 months time? > I think Erik is confusing IBM and Apple :) Apple doesn't break things to sell you a better anything. What they do is to have a real slow OS that requires serious hardware to make it perform decently. And they really ream you for the performance by charging 2 or 3 times what the hardware is worth. Evan is right, there WILL be a 25MHz 040 right away, but it will cost $20K. -- **************************************************** * I want games that look like Shadow of the Beast * * but play like Leisure Suit Larry. * ****************************************************
peter@sugar.hackercorp.com (Peter da Silva) (06/16/91)
In article <1135@stewart.UUCP> jerry@stewart.UUCP (Jerry Shekhel) writes: > Show me how the '486 is braindead, Dave. Show me that you can do more, > Dave, than post the same old ignorant Amigoid ramblings full of BS phrases > like "braindead". Well, I wouldn't call the 486 (or even the 386) braindead. I'd call them crippled, though. Even the 68000 family is short of GP registers for a '90s processor, and the 386/486 only has like 3-4 what you could call "general purpose" registers. This doesn't show up in benchmarks, generally, because they are usally either simple algorithms that don't need many registers, or feindishly complex ones that hurt the 68000 nearly as badly. But I'm sure that algorithms like BitBlt on the 386/486 require a lot of register juggling. Of course, as 90% of the 386/486 processors in the world are used, they're just faster 8088s. That *is* crippling. -- Peter da Silva. `-_-' <peter@sugar.neosoft.com>. 'U` "Have you hugged your wolf today?"
bernie@metapro.DIALix.oz.au (Bernd Felsche) (06/17/91)
In <22460@cbmvax.commodore.com> daveh@cbmvax.commodore.com (Dave Haynie) writes: >I guess Sun would be a little surprised to hear that, since they replaced the >68K with SPARC, and in fact, really hadn't quite "taken off" as a major force >in the workstation market until they did. But I do agree that the SPARC >architecture is weak when compared to the current alternatives. Clone makers >don't care about that, they're not going to go out and build a brand new >computer based on 88110 or R4000 or any merchant PA-RISC that they can >eventually do. They're going to copy something. SPARC machines are faster >than 80x86s, and Sun's making it easy for these guys to make these copies (of >course, not real fast ones, but the low end systems). What, no "real fast ones"? I hear that Cypress makes 40MHz set which run just as fast as anything "produced" by Sun and some "clone" makers actually get their systems using these chipsets running faster. There's also a 40MHz, single-chip SPARC on the horizon. That's an awful lot of grunt. Some of this discussion really should be in comp.arch, but all of these people who expound the virtues of RISC often ignore the fact that binaries on RISC are larger. As an example, I've had binaries explode by 100%, transferring from 68030 to 88000. Even with optimizers on, there was still a 50% growth. The RISC therefore needs more native MIPS to achieve the same real throughput. The 68k architecture has a fairly compact instruction set, probably because the instructions vary in in length, whereas RISC architectures tend to have fixed length instructions. It costs more to store the binaries both on disk, and while executing. Don't get me wrong, RISC has its place, I just don't think it's on the kitchen table (yet). As a low-end workstation CPU, the 68k should be more cost-effective. I simply can't imagine why Sun didn't continue with the 68k workstations at entry level. Even their last 68k series was knobbled through not making the best use that they could of the `030. Spec's were not significantly better than for the `020 predecessors. The `040 is not the beginning of the end. It's a continuation of the development. If we ever see the `050, it would have to be another quantum leap above the `040, with much extra functionality. The 030 to 040 increment incorporated the FPU, the 020 to 030 increment incorporated the MMU, the 010 to 020 brought 32 bits. What else can Motorola provide, except 64 bits? (I know that other features were also added at the various stages, I'm just mentioning the more visible.) If we don't see the `050, then the `040 has a very long product life. One never knows if CSG will start to make them if Motorola lose interest. :-) -- Bernd Felsche, _--_|\ #include <std/disclaimer.h> Metapro Systems, / sold \ Fax: +61 9 472 3337 328 Albany Highway, \_.--._/ Phone: +61 9 362 9355 Victoria Park, Western Australia v Email: bernie@metapro.DIALix.oz.au
peterk@cbmger.UUCP (Peter Kittel GERMANY) (06/17/91)
In article <1991Jun15.173342.2832@neon.Stanford.EDU>, torrie@cs.stanford.edu (Evan Torrie) writes: > > Well, DEC does have a range of high-end 80x86 PC's (interestingly, > just re-badged Intel machines). And again, Europe does it different. Here, DEC sells *Commodore MS-DOS PCs* to authorities (perhaps among other brands, don't know details). -- Best regards, Dr. Peter Kittel // E-Mail to \\ Only my personal opinions... Commodore Frankfurt, Germany \X/ {uunet|pyramid|rutgers}!cbmvax!cbmger!peterk
melling@cs.psu.edu (Michael D Mellinger) (06/18/91)
In article <1991Jun17.052423.7631@metapro.DIALix.oz.au> bernie@metapro.DIALix.oz.au (Bernd Felsche) writes:
Some of this discussion really should be in comp.arch, but all of
these people who expound the virtues of RISC often ignore the fact
that binaries on RISC are larger. As an example, I've had binaries
explode by 100%, transferring from 68030 to 88000. Even with
optimizers on, there was still a 50% growth. The RISC therefore needs
more native MIPS to achieve the same real throughput.
Disk space is becoming very cheap. The size of the binaries is
becoming less of a factor(20MB floppies are about to become popular).
As for the performance of RISC vs. CISC, take a look at the SPECmarks.
The 88K of two years ago is as fast as the 68040. Next year the new
88K will blow away the 68040(as will the R4000 in ACE machines).
As a low-end workstation CPU, the 68k should be more cost-effective. I
simply can't imagine why Sun didn't continue with the 68k workstations
at entry level. Even their last 68k series was knobbled through not
making the best use that they could of the `030. Spec's were not
significantly better than for the `020 predecessors.
Huh? By the time Moto. had the 68040 ready, Sun was ready to simply
turn up the clock speed to 40MHz and give better performance to beat
the 68040, and they don't have nearly the resources as Moto. Next
year the RISC computers take over the world. Get ready.
-Mikedaveh@cbmvax.commodore.com (Dave Haynie) (06/18/91)
In article <EACHUS.91Jun14181138@largo.mitre.org> eachus@largo.mitre.org (Robert I. Eachus) writes: >In article <22393@cbmvax.commodore.com> daveh@cbmvax.commodore.com (Dave Haynie) writes: > Anyway, looking at the Motorola track record, the '050 should be in the > "evolutionary" category, which is tracked more on architectural grounds than > performance upgrades: > 68000,68010 16 bit > 68020,68030 32 bit > 68040,68050 Cool 32 bit > I don't know what Cool 32 bit is, but I see the splits slightly >differently. > 68000, 68010: 24(27)-bit address, 16-bit data > 68020: 32(36)-bit address, 32-bit data > 68030, 68040: 32(36)-bit address, 32-bit data with cache fill support > 68050: ???-bit address, 64(128-bit) data If you see things that way, you're ignoring the underlying architecture, for other than the 68050 (which we don't know about). The 68020 and the 68030 are extremely similar. The integration of the MMU made for a lower cost and high performance (package to package delays between the '020 and '851 added a wait state to memory), but it didn't radically change any of the underlying technology. And 68030 synchronouy/burst mode as well was a very minor change, a simple enhancement of the basic 68020 asynchronous cycle (eg, you make the external logic responsible for being synchronous and you get to remove all internal synchronizers, and <poof>, it goes faster). The 68040, on the other hand, is a whole new ball of wax. The entire machine architecture was revised. Lots of instructions are hard-wired. The pipeline is amazingly more complex than on an '020/'030. There are two MMUs, versus a single one, and they are programmed differently (the '851 has a few more instructions than the '030, but they're largely compatible with one another if you stick to the common instructions; they are logically the same machine). You went from direct mapped logical caching on the '020/'030 to four set associative physical caching on the '040. Even the '040's bus has been completely revised. Really, the architectural difference between the '020/'851 and the '030 is negligable, all the performance enhancements are simple evolutionary changes (not to imply they were easy to do, but consider than it was only about a year from the time the '851 was shipping to the time we got '030 samples). The '030 to '040 change was quite revolutionary. > The 68050 is of course my WAG, but the easiest way to pump more >performance out of the architecture without increasing the clock would >be to add more data lines. (I'm sure the cache sizes will also >increase, but the remaining advantage from that is in the noise.) Of course. But simply building an external 64 bit bus or larger cache onto a 68040 architecture doesn't make an architecturally distinct processor. It could very well make a processor that goes two or three times faster, and it could make that happen much sooner than going to some totally new architecture. So that's a good thing. And it makes lots of sense, you don't necessarily even figure out how to best use a new architecture in its first incarnation. Also, it takes a Real Long Time to come up with a substantially new machine architecture. So, were I Motorola (or anyone else making similar decisions, for that matter), I would do things much the way Motorola appears to be doing them. You build a new architecture, then upgrade it. At the same time you're working on that upgrade, you're starting a really advanced architecture. That way, even though your architectural improvements take N years to achieve, you double your performance every N/2 years. -- Dave Haynie Commodore-Amiga (Amiga 3000) "The Crew That Never Rests" {uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy "This is my mistake. Let me make it good." -R.E.M.
daveh@cbmvax.commodore.com (Dave Haynie) (06/18/91)
In article <1135@stewart.UUCP> jerry@stewart.UUCP (Jerry Shekhel) writes: >daveh@cbmvax.commodore.com (Dave Haynie) writes: >>Since the leap from 68030 to 68040 was greater than the leap >>from 80386 to 80486, it sure looks like Motorola's in the lead >What the hell are you talking about? I am taking about microprocessor architecture. You are, apparently, looking a little bit at feature lists. >The "leaps" in both product lines were exactly the same (on-board FPU, True, they both have on-board FPUs with similar functionality. The '040 FPU design is more sophisticated, more like a RISC FPU, which is why its instructions execute in considerably fewer instructions. The latest external FPUs, 68882 versus 80387, were essentially on a par. Therefore, I consider Motorola to have made a much greater leap in this area. >cache, The 80486 cache is a single unified, write-through cache. A single MMU to go along with it, with a small TLB. The '040 has separate I and D caches which operate in parallel, with write-through or snoopable copyback, and large TLBs. Therefore, I consider Motorola to to have made a greater leap in this area. >Why do all of you Amigoids have it etched upon your brains that Motorola is, >was, and always will be "clearly superior"? I'm not saying anything peculiar to "Amigoid", or, in fact, anything that hasn't been said outside the "fans of Intel" segment of the industry. Intel has no choice but be "clearly inferior"; they can't change anything that will cause MS-DOS to fail, regardless of the existence of any available technology that can be easily assimilated into the Intel architecture that could speed up the operating of UNIX or OS/2. The unified cache is an obvious example of this -- Intel's even using split caches themselves in their other chips, because they simply result in better performance. >>Intel has been pushing the idea of using an i860 as a coprocessor, despite >>the fact it's not all that well suited to the job >Sorry Dave. The i860 is EXTREMELY well-suited to the job of a coprocessor, >for graphics or floating point. Sorry Jerry, it isn't. The fact that the i860's data cache operates only in copy-back mode, and provides no snooping mechanisms, make it ill suited for use as a coprocessor, since all cache consistency must be handled in software alone, that's a significant performance hit. Both of these problems are apparently cured in the new chip. >I cannot believe the ride they're offering customers now, with their >486SX/487SX processors. OK, I can accept the idea of the 486SX, for those >people who don't need the FPU or cannot afford the full 486. Yet, the 486SX, at least presently, costs the same to make as the 486, since it's the same die. They're selling, as always, based on their captive market. >But pricing the 487SX as a coprocessor (big bucks) is really sleazy, >considering it is nothing more than a full 486 repackaged with an extra pin >that turns off the 486SX. I don't see it as being any different. They're doing this one because Intel only offers single piece pricing on their coprocessors, so they make a mint on them. You don't really think a $500 '387 costs that much more to make than a $50 '882 offering Morotola systems similar FPU performance, do you? Intel just likes to sell them to end-users themselves, rather than pricing them in a range systems houses can afford to include in every unit. Strangely enough, that's spawned the creation of Weitek math chips with Intel bus pinouts. Weitek cancelled the 68030 version because nearly every '030 system ever made came with a 6888x chip (at least at the time). >MS-DOS is truly "braindead", but the '486 is not. The '486, in its native mode, isn't really that great a 32 bit processor, but it isn't bad. The architectural baggage it carries around necessary to support MS-DOS and 8086 emulation modes is what makes it ugly. The unified cache without copyback, the simplified MMU, various other architectural details make the '486 inferior to any other 32 bit processor of this generation. Which is why so many '486s are sold into PC Clones, those that go elsewhere are virtually insignificant. A few very companies pick the '486 for use outside the PC Clone market (other than Intel), and that's usually because they're running software developed on the PC. Look anywhere else, you'll see 680x0s, 29Ks, R3000s, SPARCs, maybe even an ARM or a Transputer occasionally. People use '486s because the run MS-DOS. >Show me that you can do more, Dave, than post the same old ignorant Amigoid >ramblings full of BS phrases like "braindead". Sure seems to me that you're the one not supplying any information. If I say "it is", you can sure say "it isn't". But when you ask me to supply fact to back my points without supplying any yourself, you're the one who's going to come out looking foolish. >| JERRY J. SHEKHEL | POLYGEN CORPORATION | When I was young, I had to walk | -- Dave Haynie Commodore-Amiga (Amiga 3000) "The Crew That Never Rests" {uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy "This is my mistake. Let me make it good." -R.E.M.
daveh@cbmvax.commodore.com (Dave Haynie) (06/18/91)
In article <1991Jun17.052423.7631@metapro.DIALix.oz.au> bernie@metapro.DIALix.oz.au (Bernd Felsche) writes: >In <22460@cbmvax.commodore.com> daveh@cbmvax.commodore.com (Dave Haynie) writes: >>SPARC machines are faster than 80x86s, and Sun's making it easy for these >>guys to make these copies (of course, not real fast ones, but the low end >>systems). >What, no "real fast ones"? I hear that Cypress makes 40MHz set which >run just as fast as anything "produced" by Sun and some "clone" makers >actually get their systems using these chipsets running faster. Well, it's easy to get fast SPARC Integer processors. Entire chip sets as good or better than the ones Sun's using in the SparcStation 2 are announced but not yet shipping in volume, at last last I heard (a few weeks ago in "Microprocessor Report"). The off-the-shelf parts from LSI are slower versions of the Sun chip set. >There's also a 40MHz, single-chip SPARC on the horizon. That's an >awful lot of grunt. Yeah, a year or two ago I figured SPARCs, along with most of the other RISCy toys, would be getting to the integration level of 68040 and 80486 pretty quickly. That's where the smaller IPU and FPU really pay off, you'll fit more cache or alternate processing units in the same silicon space, or the same in less (going for cheap rather than fast). >Some of this discussion really should be in comp.arch, but all of >these people who expound the virtues of RISC often ignore the fact >that binaries on RISC are larger. As an example, I've had binaries >explode by 100%, transferring from 68030 to 88000. Lots of it depends on what you're doing, but in general, you would expect at least some code increase. First of all, all the real RISCs out there have very orthogonal 32 bit long instructions, rather than these variable length operataions on CISCs and Transputers (now there's some real instruction reduction, though you need zillions to get any work done). You lose alot on complex things with memory addressing and preinc/postdec thrown in, though most RISCs win in math with three operand instructions. >Even with optimizers on, there was still a 50% growth. The RISC therefore needs >more native MIPS to achieve the same real throughput. True. But even Marketing departments are getting away from quoting native MIPS these days. In fact, the only one I've seen lately doing that is INMOS/ Thompson for their Transputers. >It costs more to store the binaries both on disk, and while executing. It costs more to store them, but less to decode them. With the relative priorities these days between speed and the need to save memory, I think the RISC machines are on the right track. >As a low-end workstation CPU, the 68k should be more cost-effective. I >simply can't imagine why Sun didn't continue with the 68k workstations >at entry level. Like HP did, for instance. My guess is that they wanted to push SPARC with all their effort, to make sure that none of the IC houses had any questions about their commitment. And with all the clone-ability, they're definitely trying to move upscale. >Even their last 68k series was knobbled through not making the best use that >they could of the `030. Yeah, I think they pretty much just dropped the '030 into their '020 systems, like Apple did with the first '030 based Mac IIs. HP and Apollo did it better. >One never knows if CSG will start to make them if Motorola lose >interest. :-) If only they had a 0.8um CMOS line.... -- Dave Haynie Commodore-Amiga (Amiga 3000) "The Crew That Never Rests" {uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy "This is my mistake. Let me make it good." -R.E.M.
chucks@pnet51.orb.mn.org (Erik Funkenbusch) (06/19/91)
mykes@amiga0.SF-Bay.ORG (Mike Schwartz) writes: >In article <1991Jun15.173528.3156@neon.Stanford.EDU> torrie@cs.stanford.edu (Evan Torrie) writes: >>chucks@pnet51.orb.mn.org (Erik Funkenbusch) writes: >> >>>Apple is NOTORIOUS for breaking things so that they can sell you a better >>>version later. this is why i believe the first 040 macs will be 16 Mhz. so >>>they can sell you a 25 Mhz later.. >> >> Can I quote you on this in about 4 months time? >> > >I think Erik is confusing IBM and Apple :) Apple doesn't break things to >sell you a better anything. What they do is to have a real slow OS that >requires serious hardware to make it perform decently. And they really >ream you for the performance by charging 2 or 3 times what the hardware >is worth. Oh really? look at apple's track record. Mac II (ok, so it was an entry level machine, apples first attempt at 32 bit.) it was a 16 Mhz 020. then came the speedy Mac IIx (a 16Mhz 030.. wow.. speedy), then the IIcx (almost a year later, still a 16 Mhz machine) then FINALLY comes the IIci which is 25 Mhz, the speed at which their machines SHOULD have been clocked for to begin with). then comes the IIfx which uses a 50 Mhz 030 chip clocked at 40 Mhz(!) boy, that's really squeezing the performance.. in each situation apple has PROVEN that they come out with an inferior product to what should have been designed, and when sales slack, they pump up the power. (forcing you to continually upgrade if you want the best performance). is it any wonder why i believe the first 040 mac will be 16 Mhz? > >Evan is right, there WILL be a 25MHz 040 right away, but it will cost >$20K. > >-- >**************************************************** >* I want games that look like Shadow of the Beast * >* but play like Leisure Suit Larry. * >**************************************************** .--------------------------------------------------------------------------. | UUCP: {amdahl!tcnet, crash}!orbit!pnet51!chucks | "I know he's come back | | ARPA: crash!orbit!pnet51!chucks@nosc.mil | from the dead, but do | | INET: chucks@pnet51.orb.mn.org | you really think he's | |-------------------------------------------------| moved back in?" | | Amiga programmer at large, employment options | Lou Diamond Philips in | | welcome, inquire within. | "The First Power". | `--------------------------------------------------------------------------'
awessels@ccwf.cc.utexas.edu (Allen Wessels) (06/20/91)
In article <5171@orbit.cts.com> chucks@pnet51.orb.mn.org (Erik Funkenbusch) writes: >year later, still a 16 Mhz machine) then FINALLY comes the IIci which is 25 >Mhz, the speed at which their machines SHOULD have been clocked for to begin >with). then comes the IIfx which uses a 50 Mhz 030 chip clocked at 40 Mhz(!) Really? Were those 25Mhz chips being produced in the quantities Apple would need? >designed, and when sales slack, they pump up the power. (forcing you to >continually upgrade if you want the best performance). is it any wonder why i >believe the first 040 mac will be 16 Mhz? Hmm, sound like that same argument could be applied to the Amiga. 3rd parties almost always have a faster processor available for a given machine. It isn't any wonder to me that you think the first Apple-produced 040 Mac would be a 16Mhz machine. If I didn't know anything about a machine or the company that produced it, or the business realities of the situation, I could imagine all sorts of absurd things as well.
eachus@largo.mitre.org (Robert I. Eachus) (06/20/91)
In article <22515@cbmvax.commodore.com> daveh@cbmvax.commodore.com (Dave Haynie) writes: In article <EACHUS.91Jun14181138@largo.mitre.org> eachus@largo.mitre.org (Robert I. Eachus) writes: > I don't know what Cool 32 bit is, but I see the splits slightly >differently. > 68000, 68010: 24(27)-bit address, 16-bit data > 68020: 32(36)-bit address, 32-bit data > 68030, 68040: 32(36)-bit address, 32-bit data with cache fill support > 68050: ???-bit address, 64(128-bit) data If you see things that way, you're ignoring the underlying architecture, for other than the 68050 (which we don't know about).... Motorola did a hell of a lot of things right in the 68040, and you are correct in pointing out that under the skin, the 68040 looks more like an 88000 that a 68020. My grouping of the 68030 and 68040 in that list was NOT intended to imply that the innards were at all related, just that memory designs that support the 68030 do a reasonably good job of supporting the 68040, while both you and I (come on, admit it Dave :-) switched bus designs completely between the 68020 and the 68030. Getting maximum performance out of the 68020 meant an external cache, while low end machines used the on-board cache. The number of clocks needed to feed data to a 68020 meant that any simple 32-bit data bus could make it happy. While it is possible to treat a 68030 like an overgrown 68020, and Apple did so in some of their machines, with the 68030 you had to improve the memory system (and the bus) to keep up with the chip which could now eat data (or instructions) at more than twice the previous rate. The 68040 can draw from main memory a little faster than the 68030, but the cache is much more effective, so a bus designed for a 68030 at the same clock speed can feed a 68040 (modulo that one clock cycle), just like the same bus bandwidth would feed a 68000 and 68010. My guess is that the 68030 and 68040 are eating all the data a 32-bit wide bus can handle so the 68050 will have to go wider, and it will be time to redesign systems again, not just CPU boards. (Of course, in the meantime some people will pay good money for 40 and 50 MHz 68040 systems, and feeding those will be a real challenge!) -- Robert I. Eachus with STANDARD_DISCLAIMER; use STANDARD_DISCLAIMER; function MESSAGE (TEXT: in CLEVER_IDEAS) return BETTER_IDEAS is...
torrie@cs.stanford.edu (Evan Torrie) (06/20/91)
mykes@amiga0.SF-Bay.ORG (Mike Schwartz) writes: >In article <1991Jun15.173528.3156@neon.Stanford.EDU> torrie@cs.stanford.edu (Evan Torrie) writes: >>chucks@pnet51.orb.mn.org (Erik Funkenbusch) writes: >> >>>Apple is NOTORIOUS for breaking things so that they can sell you a better >>>version later. this is why i believe the first 040 macs will be 16 Mhz. so >>>they can sell you a 25 Mhz later.. >> >> Can I quote you on this in about 4 months time? >> >Evan is right, there WILL be a 25MHz 040 right away, but it will cost >$20K. Can I quote you on this in about 4 months time? :-) -- ------------------------------------------------------------------------------ Evan Torrie. Stanford University, Class of 199? torrie@cs.stanford.edu "Lay me place and bake me pie, I'm starving for me gravy... Leave my shoes and door unlocked, I might just slip away - hey - just for the day."
merlin@presto.UUCP (Jeff W. Hyche) (06/20/91)
In article <!47H$dhy@cs.psu.edu> melling@cs.psu.edu (Michael D Mellinger) writes: > > >Huh? By the time Moto. had the 68040 ready, Sun was ready to simply >turn up the clock speed to 40MHz and give better performance to beat >the 68040, and they don't have nearly the resources as Moto. Next >year the RISC computers take over the world. Get ready. > Some how I don't see this happening. I would think there will be a merge the the technologys. -- // Jeff Hyche There can be only one! \\ // Usenet: hychejw@infonode.ingr.com \X/ UUCP: ...!uunet!sci34hub!presto!merlin
dant@ryptyde.UUCP (Daniel Tracy) (06/20/91)
Responding to the following: "then comes the IIfx which uses a 50 Mhz 030 chip clocked at 40 Mhz(!) boy, that's really squeezing the performance.. in each situation apple has PROVEN that they come out with an inferior product to what should have been designed, and when sales slack, they pump up the power. (forcing you to continually upgrade if you want the best performance). is it any wonder why i believe the first 040 mac will be 16 Mhz?" Are you kidding? You're saying something to the effect that, in the IIfx, a 50MHz 030 is used, but it is slowed down to 40MHz for some reason/somehow? Please explain this, this is very interesting.
navas@cory.Berkeley.EDU (David C. Navas) (06/21/91)
In article <75@ryptyde.UUCP> dant@ryptyde.UUCP (Daniel Tracy) writes: >Are you kidding? You're saying something to the effect that, in the IIfx, >a 50MHz 030 is used, but it is slowed down to 40MHz for some reason/somehow? >Please explain this, this is very interesting. I believe a sort of sleight ot hand was used. 40Mhz 68030s are actually 'failed' 50Mhz 68030's. At least that's the story as I heard it. Seems reasonable to assume that as Motorola's "success rate" goes up, 40Mhz become rarer, and you *might* start using 50Mhz parts.... And oh, by the way, memory allocation routines: AllocAbs(), Allocate(), AllocEntry(), AllocMem(), AllocVec() [2.0], AllocRemember(), etc. Hmm, I guess we do have "Dynamic Memory Allocation". Whatta you think? :) David Navas navas@cory.berkeley.edu 2.0 :: "You can't have your cake and eat it too." Also try c186br@holden, c260-ay@ara and c184-ap@torus
jerry@polygen.uucp (Jerry Shekhel) (06/21/91)
daveh@cbmvax.commodore.com (Dave Haynie) writes: > >The unified cache is an obvious example >of this -- Intel's even using split caches themselves in their other chips, >because they simply result in better performance. > Are you suggesting that the unified cache is a limitation designed to ensure MS-DOS compatibility? How would a split cache break MS-DOS? > >Sorry Jerry, it isn't. The fact that the i860's data cache operates only in >copy-back mode, and provides no snooping mechanisms, make it ill suited for >use as a coprocessor, since all cache consistency must be handled in software >alone, that's a significant performance hit. Both of these problems are >apparently cured in the new chip. > And yet the i860 is used all over the place as a coprocessor, but only in a couple of places as a CPU! I don't know enough about it, but it makes me wonder. > >Yet, the 486SX, at least presently, costs the same to make as the 486, since >it's the same die. They're selling, as always, based on their captive market. > I agree. Actually, I'd bet that the 486SX costs less to produce, since the yield is probably much greater. I wouldn't be surprised if a large part of 486SX production consisted of reject 486 chips that failed the speed test or ended up with defective FPUs. > >The '486, in its native mode, isn't really that great a 32 bit processor, but >it isn't bad. The architectural baggage it carries around necessary to support >MS-DOS and 8086 emulation modes is what makes it ugly. The unified cache >without copyback, the simplified MMU, various other architectural details make >the '486 inferior to any other 32 bit processor of this generation. Which is >why so many '486s are sold into PC Clones, those that go elsewhere are virtlly >insignificant. A few very companies pick the '486 for use outside the PC Clone >market (other than Intel), and that's usually because they're running software >developed on the PC. > While I agree that the 486 instruction set is a bit dated, I wonder what you mean by "simplified MMU". It does the job. How is it "simplified" or "limited"? Also, there are several companies which produce multiprocessor 486 systems that have nothing to do with DOS. As for me, I got a 386 because it's a cheap UNIX platform that's plenty fast enough for me. Once you have UNIX, who cares what the processor is? I used an SGI for years before finding out what kind of CPU it had. > >Sure seems to me that you're the one not supplying any information. If I >say "it is", you can sure say "it isn't". But when you ask me to supply fact >to back my points without supplying any yourself, you're the one who's going >to come out looking foolish. > I don't think so, Dave. You said, "It sucks because it's braindead." I said, "Prove it." How was my reply foolish? Maybe your original statement was the more foolish one? >Dave Haynie Commodore-Amiga (Amiga 3000) "The Crew That Never Rests" -- +-------------------+----------------------+---------------------------------+ | JERRY J. SHEKHEL | POLYGEN CORPORATION | When I was young, I had to walk | | Drummers do it... | Waltham, MA USA | to school and back every day -- | | ... In rhythm! | (617) 890-2175 | 20 miles, uphill both ways. | +-------------------+----------------------+---------------------------------+ | ...! [ princeton mit-eddie bu sunne ] !polygen!jerry | | jerry@polygen.com | +----------------------------------------------------------------------------+
torrie@cs.stanford.edu (Evan Torrie) (06/22/91)
jerry@polygen.uucp (Jerry Shekhel) writes: >daveh@cbmvax.commodore.com (Dave Haynie) writes: >> >>The unified cache is an obvious example >>of this -- Intel's even using split caches themselves in their other chips, >>because they simply result in better performance. >> >Are you suggesting that the unified cache is a limitation designed to ensure >MS-DOS compatibility? How would a split cache break MS-DOS? MS-DOS used to [still does?], and various MS-DOS-based programs, make use of self-modifying code. In a split cache environment, any self-modifying code immediately falls over in a screaming heap, because any "code" modifications are reflected only in the data cache, not in the instruction cache. In my opinion, this was evidently a major, major compromise on the part of Intel to accomodate backward compatibility with MS-DOS, since as Dave notes, Intel's other microprocessors [along with Motorola and the RISC manufacturers] use split I/D caches. >> >>Yet, the 486SX, at least presently, costs the same to make as the 486, since >>it's the same die. They're selling, as always, based on their captive market. >> >I agree. Actually, I'd bet that the 486SX costs less to produce, since the >yield is probably much greater. I wouldn't be surprised if a large part of >486SX production consisted of reject 486 chips that failed the speed test or >ended up with defective FPUs. According to EE Times, _all_ of the 486SX production consists of real 486 chips with the FPU disabled. (Apparently, a real 486SX die won't be fabricated until the end of the year). In the meantime, you could also buy a 487SX, which just happens to be yet another real 486, this time with everything BUT the FPU disabled. So for more than the price of a single 486, you get to buy TWO castrated 486s, which together run slower than the single 486. Is it any wonder Intel is despised by so many?? -- ------------------------------------------------------------------------------ Evan Torrie. Stanford University, Class of 199? torrie@cs.stanford.edu "Lay me place and bake me pie, I'm starving for me gravy... Leave my shoes and door unlocked, I might just slip away - hey - just for the day."
peter@Sugar.NeoSoft.com (Peter da Silva) (06/23/91)
In article <1991Jun22.065702.24890@neon.Stanford.EDU> torrie@cs.stanford.edu (Evan Torrie) writes: > In the meantime, you could also > buy a 487SX, which just happens to be yet another real 486, this time with > everything BUT the FPU disabled. No, it's even worse. The 487SX is a 486 with an extra pin that disables the 486SX and replaces it. The 486SX is just a dongle when you have the 487SX installed. Yes, we dislike Apple for using legal shenanigans to maximise profits at the expense of the entire software industry... but what IBM and Intel do makes them almost look like the good guys. -- Peter da Silva. `-_-' <peter@sugar.neosoft.com>. 'U` "Have you hugged your wolf today?"
chucks@pnet51.orb.mn.org (Erik Funkenbusch) (06/26/91)
awessels@ccwf.cc.utexas.edu (Allen Wessels) writes: >In article <5171@orbit.cts.com> chucks@pnet51.orb.mn.org (Erik Funkenbusch) writes: > >>year later, still a 16 Mhz machine) then FINALLY comes the IIci which is 25 >>Mhz, the speed at which their machines SHOULD have been clocked for to begin >>with). then comes the IIfx which uses a 50 Mhz 030 chip clocked at 40 Mhz(!) > >Really? Were those 25Mhz chips being produced in the quantities Apple would >need? Yes, Motorola has been producing 25 Mhz 030's in great quanitites for years.. I believe that about the time the IIcx came out There were 25 Mhz NeXT's, 25Mhz Amiga's etc.. Apples sales of the IIcx at the time they created it weren't increadible either. it became a better seller later on. > >>designed, and when sales slack, they pump up the power. (forcing you to >>continually upgrade if you want the best performance). is it any wonder why i >>believe the first 040 mac will be 16 Mhz? > >Hmm, sound like that same argument could be applied to the Amiga. 3rd >parties almost always have a faster processor available for a given machine. But Commodore's Machines are clocked at the Maximum rating for their Chips. you didn't explain why the IIfx is only clocked at 40 Mhz, when the chip is rated at 50 Mhz. CBM releases what's the best at the time. I'm sure they could have released a 50Mhz 3000, but what's the point? when the 040 was right around the corner, they chose to reduce costs by using a cheaper chip. the IIfx uses that expensive chip but doesn't do it justice. > >It isn't any wonder to me that you think the first Apple-produced 040 Mac would >be a 16Mhz machine. If I didn't know anything about a machine or the company >that produced it, or the business realities of the situation, I could imagine >all sorts of absurd things as well. So enlighten me. .--------------------------------------------------------------------------. | UUCP: {amdahl!tcnet, crash}!orbit!pnet51!chucks | "I know he's come back | | ARPA: crash!orbit!pnet51!chucks@nosc.mil | from the dead, but do | | INET: chucks@pnet51.orb.mn.org | you really think he's | |-------------------------------------------------| moved back in?" | | Amiga programmer at large, employment options | Lou Diamond Philips in | | welcome, inquire within. | "The First Power". | `--------------------------------------------------------------------------'
daveh@cbmvax.commodore.com (Dave Haynie) (06/28/91)
In article <1147@stewart.UUCP> jerry@stewart.UUCP (Jerry Shekhel) writes: >daveh@cbmvax.commodore.com (Dave Haynie) writes: >>The unified cache is an obvious example >>of this -- Intel's even using split caches themselves in their other chips, >>because they simply result in better performance. >Are you suggesting that the unified cache is a limitation designed to ensure >MS-DOS compatibility? How would a split cache break MS-DOS? MS-DOS programs run self-modifying code out the wazoo. Separate I/D caches don't support this at all. Unified caches have no choice, of course, since they have no concept of what is data and what is program. >>Sorry Jerry, it isn't. The fact that the i860's data cache operates only in >>copy-back mode, and provides no snooping mechanisms, make it ill suited for >>use as a coprocessor, since all cache consistency must be handled in software >>alone, that's a significant performance hit. Both of these problems are >>apparently cured in the new chip. >And yet the i860 is used all over the place as a coprocessor, but only in a >couple of places as a CPU! I don't know enough about it, but it makes me >wonder. Actually, they rarely use it as a "coprocessor". More like a rather loosely coupled graphics engine, same basic function as a TI340x0, DSP, etc. In such an architecture, the i860 is restricted to a small chunk of memory which can be inefficiently shared, at least to some extent, by the host CPU. Sure, it works, you can build such a system with absolutely any processor. The C= A2232 card does this with a 4502 (6502 compatible), the BridgeCards of course work similarly with 8088/80286 processors. None of those were designed to work as coprocessors, and would make really bad tightly coupled processors. But in limited situations, anything can be used. That doesn't imply that the thing being used was designed to behave efficiently as a slave processor. >While I agree that the 486 instruction set is a bit dated, I wonder what you >mean by "simplified MMU". It does the job. How is it "simplified" or >"limited"? Simplified, in that its page table strategy requires more memory and more time to service than that of the 68040. Simplified in that its ATC is both smaller and less efficient that those of the 68040 (4 set rather than fully associative, as I recall). Simplified, in that there is no block translation feature, required for efficient mapping of things like display buffers without taking constant ATC hits. The list can go on. There is no secret here. Intel designers are certainly as up to date on the lastest computer architecture concepts as those at Motorola. The problem was simply that they had to leave out features because of all the architectural baggage they carry around supporting CPU 8086 "mode". There's only so much silicon available, and Intel had their hands tied in a way Motorola didn't. >I don't think so, Dave. You said, "It sucks because it's braindead." I said, >"Prove it." How was my reply foolish? Maybe your original statement was the >more foolish one? I said "it's bad". Sure, I didn't elaborate, this isn't comp.sys.amiga.hardware or comp.arch, and I see that after replying with backing proof, you still didn't understand the architecture you're arguing for well enough to appreciate my position. You said "no it isn't". That's a fair exchange with my unsupported argument. But you went on to ask for proof, without supplying any on your side of the argument. You still have yet to provide any. That's what I call a foolish argument, you can't win an debate by avoiding the issues. Unless maybe you're entering politics. -- Dave Haynie Commodore-Amiga (Amiga 3000) "The Crew That Never Rests" {uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy "This is my mistake. Let me make it good." -R.E.M.
daveh@cbmvax.commodore.com (Dave Haynie) (06/28/91)
In article <5216@orbit.cts.com> chucks@pnet51.orb.mn.org (Erik Funkenbusch) writes: >awessels@ccwf.cc.utexas.edu (Allen Wessels) writes: >>In article <5171@orbit.cts.com> chucks@pnet51.orb.mn.org (Erik Funkenbusch) writes: >>>year later, still a 16 Mhz machine) then FINALLY comes the IIci which is 25 >>>Mhz, the speed at which their machines SHOULD have been clocked for to begin >>>with). then comes the IIfx which uses a 50 Mhz 030 chip clocked at 40 Mhz(!) There really is a 40MHz 68030, though the 50MHz came first. The early '030 design only worked up to 33MHz; a design tweak resulted in 50MHz. But, as always, they don't get every chip made to pass at 50MHz, so they created the 40MHz part. The 40MHz CPU-Cache/20MHz system bus make sense when you consider the Mac's NuBus. Since NuBus is a synchronous bus with a relatively slow clock (10MHz), you'll die in synchronization delays if your CPU isn't in sync with it (like Mac II, IIx, IIcx, and IIci). The IIfx, more than likely, is designed with a fully synchronous motherboard<->NuBus interface. >>Really? Were those 25Mhz chips being produced in the quantities Apple would >>need? 25MHz '030s were sampled early in '88. The A2500/20 came out in the fall of '88, the A2500/30 in the fall of '89. The first 25MHz A2630 was running in prototype form at the Washington D.C. Amiga Developer's Conference in 1988. I don't recall when the IIci first shipped. -- Dave Haynie Commodore-Amiga (Amiga 3000) "The Crew That Never Rests" {uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy "This is my mistake. Let me make it good." -R.E.M.
torrie@cs.stanford.edu (Evan Torrie) (07/01/91)
daveh@cbmvax.commodore.com (Dave Haynie) writes: >25MHz '030s were sampled early in '88. The A2500/20 came out in the fall of >'88, the A2500/30 in the fall of '89. The first 25MHz A2630 was running in >prototype form at the Washington D.C. Amiga Developer's Conference in 1988. I >don't recall when the IIci first shipped. The IIci was announced September 20, 1989. -- ------------------------------------------------------------------------------ Evan Torrie. Stanford University, Class of 199? torrie@cs.stanford.edu "Cold is God's way of telling us to burn more Catholics" - Lady Whiteadder