gillies@p.cs.uiuc.edu (10/20/89)
About being "fed up with MIPS"
Two years ago I bought a Mac II, which was fast as blazes at WYSWYG
word processing. It blew the doors off the old DLION I used to use.
It compiled at 45,000 lines per minute. To friends, I predicted the
hunger for PC cycles would drop in the next five years. Most people
wouldn't need more than 3 MIPs.
But today, a 1000*1000*24bit graphics display takes at least 10 MIPS
to refresh interactively. Today's Mac II struggles at 3-4 MIPS and
1/4 the resolution. It is now "the premier platform for photographic
visual arts."
But most office clerical work needs no more than an XT 8088 machine
and WordPerfect. They don't need color, SPICE, or a large database.
Let's hope these clericals are doomed, or the PC industry will die as
they flourish.
To keep the PC industry thriving we must produce many more technicians
and design engineers, to use high-MIPS CPUs in design optimization and
simulation. There is a big possibility that the U.S. educational
system will let us down. Science and Engineering are unpopular
subjects; everyone wants to be a business major.
So I'm again predicting doom for the PC industry, but this time for
different reasons.
Don Gillies, Dept. of Computer Science, University of Illinois
1304 W. Springfield, Urbana, Ill 61801
ARPA: gillies@cs.uiuc.edu UUCP: {uunet,harvard}!uiucdcs!gilliesdmocsny@uceng.UC.EDU (daniel mocsny) (10/20/89)
In article <76700077@p.cs.uiuc.edu>, gillies@p.cs.uiuc.edu writes: > But most office clerical work needs no more than an XT 8088 machine > and WordPerfect. They don't need color, SPICE, or a large database. > Let's hope these clericals are doomed, or the PC industry will die as > they flourish. > > To keep the PC industry thriving we must produce many more technicians > and design engineers, to use high-MIPS CPUs in design optimization and > simulation. There is a big possibility that the U.S. educational > system will let us down. How interesting, to suggest adjusting US educational priorities to insure a stable market for computers. Note that I share your concern about our assuredly dangerous shift away from technical education at the same time the need for it grows. However, I have to be amused at the concept of the computer industry cloaking itself in the mantle that GM once reserved for itself, to whit: "What is good for GM is good for the country..." :-) Most market-driven enterprises attempt to adjust their operations to accommodate the realities of the market, rather than adjusting the market to fit their own perception of reality. The latter is, of course, the province of the oligopoly. Having grown large enough to largely free itself from the disturbing effects of competition and consumer opinion, the oligopoly applies its considerable resources to the important task of social engineering (or managing demand). The oligopoly seeks to bring about massive changes in the way people go about their lives, to insure a continuing need for its product. The automobile industry in the USA has enjoyed unequaled success in this. The computer industry can learn from the automobile industry. The conputer industry should not have to kill as many people or cause as much social distruption to realize similar success, since computer technology is by itself inherently benign. However, the computer industry must grasp several analogies: 1. the analogy between freeways and public-access computer networks, and the indispensable role of government in providing infrastructure on this scale; 2. how an isolated computer is like a stationary engine used, for example, for pumping water; 3. how a widely-networked computer is like an engine mounted on a wheeled vehicle, able to take its owner anywhere that time, fuel, and paved infrastructure permit; 4. how new technology requires new infrastructure, and must compete with old technology and old infrastucture: the motor industry had to specifically target and undermine the rail industry, and now the computer industry must do much the same to the motor industry. Consider the analogy (2). A computer performing some well-defined, simple task dealing only with local information is like an engine that sits in one place and does one thing, for example pump water. Since demand for water has some upper limit, once you pump that much, you don't need a bigger engine. Now consider analogy (3). Give the machine a vast world to move around in, and the user will spontaneously discover endless new things to use it for. So if anyone wants my prescription to keep MIPS demand high, I say a big chunk of the future is in telecommunications. PROVIDED that the industry can clean up its act enough to make sending information between two computers as simple as establishing a connection between two telephones, then the demand for computer power will have no upper bound. This is the simple consequence of Parkinson's Law applied to computers. Parkinson noted that bureaucracies tend to grow by a constant percentage every year regardless of whether they have any work to do, since all the employees in the bureau generate work for each other to do. Everyone has to read each other's minutes, negotiate their differences, listen to each other's complaints, overcome personality clashes, etc. Thus they each feel overburdened and pressure for staff increases. Staff increases accomplish nothing, of course, except to give the existing employees more people to spend their time interacting with. Since two computers must perform a considerable amount of work to talk with each other, the analogy is obvious, with the important exception that computers occasionally do useful work. Consider how the computer industry is now ceding several $ billions in demand for telecom to the FAX market. FAX technology is inferior to existing e-mail technology in almost every respect, except one: ease of use. As Peter F. Drucker says, the world will never lack for uneducated and incompetent people. By failing to make your product accessible to them, you limit your market to a minor niche. The 8088-PC-bound desk clerk languishes in a mostly isolated world. Much of this has to do with the impedance mismatch between the computer and it communication channel with the outside world. The mismatch is between the speed of light and the speed of paper. Essentially all the information entering and leaving the typical small office moves either on paper or via the spoken word. Obviously, requiring a human clerk to sit at either end of the communication channel and translate things for the computer puts a real limit on how much work the computer can do. To escape from this Curse of the Pharoes (i.e., papyrus), the computer industry must understand this principle: Using computers to generate paper faster is like strapping a jet engine to a horse. Sure, you might gallop a little faster that way, but a bigger engine isn't going to win the race. When a new power source shows up, you have to re-think your entire concept of "vehicle." 'Nuff said. Dan Mocsny dmocsny@uceng.uc.edu
roelof@idca.tds.PHILIPS.nl (R. Vuurboom) (10/20/89)
In article <76700077@p.cs.uiuc.edu> gillies@p.cs.uiuc.edu writes: > >About being "fed up with MIPS" > >But most office clerical work needs no more than an XT 8088 machine >and WordPerfect. They don't need color, SPICE, or a large database. >Let's hope these clericals are doomed, or the PC industry will die as >they flourish. > The PC industry today (and that of a large part of the computer industry also) has to my mind a strong analogy with that of the automobile industry in the fifties where increased engine performance was a dominating technological driving force or to go back even further to bridge building before the turn of the century when tensile strength (of the steel alloys) also had a dominating effect on bridge design. There are two points to note about this: 1. In general in any particular engineering discipline one particular attribute dominates development only for so long after a while other attributes become dominant. 2. The search for improvement in that particular attribute will continue but will no longer be leading in that discipline. Racing cars today no longer represent the forefront of passenger car technology but have become a separate technology. Tensile strength research is still important but no longer in relation to bridge building. If history repeats itself (and its up to you to decide if it does or does'nt) then I think one can make the following predictions: 1. A point will come when higher cpu performance (in the traditional "more mips" sense) will no longer be the dominating trend in computer design. 2. Higher performant computers (in the traditional sense of "more mips") will always be built but at a point in time (the same point as above ? :-) they will no longer represent the forefront of "normal" computer technology but will have become a separate species (the supercomputer of today will not be the laptop of tomorrow...:-). Theres another analogy with the automobile industry and thats the rate at which new models were introduced...in the seventies the rate of model introduction was reduced (to me the sign of a mature technology). This (I assume) will also occur in the computer industry at a "certain point in time". -- Artificial Intelligence: When computers start selling stocks because its Friday the 13th.... Roelof Vuurboom SSP/V3 Philips TDS Apeldoorn, The Netherlands +31 55 432226 domain: roelof@idca.tds.philips.nl uucp: ...!mcvax!philapd!roelof
bzs@world.std.com (Barry Shein) (10/22/89)
> To keep the PC industry thriving we must produce many more technicians > and design engineers, to use high-MIPS CPUs in design optimization and > simulation. There is a big possibility that the U.S. educational > system will let us down. The U.S. Educational System will let us down??? How about the industries who supposedly need these people? It never ceases to amuse me how industry types will wring their hands bitching that universities aren't teaching this or that. Yet, when it comes time to make a donation, they're nowhere to be found or do palliative things that look good in the newspaper (donate a few PC's) but balk at doing anything substantive (real $$ to develop real programs and hire real teaching and support staff.) It's only important to them when they have the podium, not the checkbook. If US Industry has a need of academia they can put their money where their collective mouths are. Otherwise they should take what they get. I doubt you could even *find* a university that, if offered sufficient funding, wouldn't develop an educational program tailored to whatever your little hearts desire. Go visit some computer science departments, other than the big four or five, and watch them scraping around for even the most basic teaching facilities, fighting with the administration to just get a few dumb CRT's and (much more importantly) unable to pay salaries which will draw anyone beyond re-tread mathematicians who could care less about actually using computers (you can spot them easily, they refer to programmers as "button-pushers" and sneer, it's not unusual to see cs depts with 3/4 logicians and the like cuz anyone who can actually do anything practical is out doing it for a real salary and real support for their work.) When industry does show up with a supposed donation either they have ulterior motives (let's donate one of our XY1000's so they'll buy more of them!) or they're looking to get some cheap programming done under the guise of "research" (with a tax deduction to boot!) Shit, I've sat in any number of meetings where certain big computer companies stood up and said: "We'd love to support RESEARCH here, we need proposals that CAN BE PRODUCTIZED WITHIN TWO YEARS." What about education? Nervous silence, er, how do we PRODUCTIZE education??? Get serious. There's occasional generosity, it seems to come and go with changes in the corporate tax laws. And it's almost always limited to donations of some equipment which ain't enough to put good teachers in front of classrooms. Anyhow, if you're not part of the solution you're part of the problem and all that. -- -Barry Shein Software Tool & Die, Purveyors to the Trade | bzs@world.std.com 1330 Beacon St, Brookline, MA 02146, (617) 739-0202 | {xylogics,uunet}world!bzs
peter@ficc.uu.net (Peter da Silva) (10/22/89)
In article <2526@uceng.UC.EDU> dmocsny@uceng.UC.EDU (daniel mocsny) writes: > 2. how an isolated computer is like a stationary engine used, for > example, for pumping water; A-men. > The 8088-PC-bound desk clerk languishes in a mostly isolated world. I've been saying this for years. > Using computers to generate paper faster is like strapping a jet > engine to a horse. A large part of the problem is that the operating system on this 8088-based PC was designed to allow one person to use the computer at a time. A network is a multi-user environment, and without a multi-user computer (multitasking with at least application-level protection) you can't make effective use of one. -- Peter da Silva, *NIX support guy @ Ferranti International Controls Corporation. Biz: peter@ficc.uu.net, +1 713 274 5180. Fun: peter@sugar.hackercorp.com. `-_-' "ERROR: trust not in UUCP routing tables" 'U` -- MAILER-DAEMON@mcsun.EU.net
ching@pepsi.amd.com (Mike Ching) (10/23/89)
In article <1989Oct22.001109.1008@world.std.com> bzs@world.std.com (Barry Shein) writes: > >> To keep the PC industry thriving we must produce many more technicians >> and design engineers, to use high-MIPS CPUs in design optimization and >> simulation. There is a big possibility that the U.S. educational >> system will let us down. > >The U.S. Educational System will let us down??? > >How about the industries who supposedly need these people? > >It never ceases to amuse me how industry types will wring their hands >bitching that universities aren't teaching this or that. > >Yet, when it comes time to make a donation, they're nowhere to be >found or do palliative things that look good in the newspaper (donate >a few PC's) but balk at doing anything substantive (real $$ to develop >real programs and hire real teaching and support staff.) > [verbiage deleted] > >When industry does show up with a supposed donation either they have >ulterior motives (let's donate one of our XY1000's so they'll buy more >of them!) or they're looking to get some cheap programming done under >the guise of "research" (with a tax deduction to boot!) First you say that industry would get what they need from education if they donated funds and then you say that getting what they need is an ulterior motive. Were you just venting or did you have a point? mike ching
cm@yarra.oz.au (Charles Meo) (10/23/89)
> In article <76700077@p.cs.uiuc.edu> gillies@p.cs.uiuc.edu writes: > > >But most office clerical work needs no more than an XT 8088 machine >and WordPerfect. They don't need color, SPICE, or a large database. >Let's hope these clericals are doomed, or the PC industry will die as >they flourish. I am a little puzzled by this attitude. Why should the 'clericals' disappear just because they don't need more computing power than they have? It seems axiomatic to me that there is never a need to apply more force (read mips) than is necessary to do a job. Others have written of parallels with the US auto industry: I'd agree with that! There are lots of salesmen and technocrats trying to sell you a muscle car when you might only need something to get to the local shops and back... why have a Stingray when (a) you hate driving and (b) a Honda Z will do, and is cheaper. Business is about satisfying customer needs/wants. The computer industry as whole should be as concerned with providing low-power users with the resources they require, as with providing heavy-duty iron for big applications: always provided there is a buck in it! I would discount the possibility that large numbers of satisfied users happily running with PCXT will be a millstone aroud the neck of R&D. All this means is that the demand for hot PC's will be less, and hence costs of making them will settle at a higher point. They don't stop developing fast cars just because not many people buy them! chuck -- 4GL: crushing snails with sledgehammers
davidsen@crdos1.crd.ge.COM (Wm E Davidsen Jr) (10/23/89)
In article <76700077@p.cs.uiuc.edu>, gillies@p.cs.uiuc.edu writes: | But most office clerical work needs no more than an XT 8088 machine | and WordPerfect. They don't need color, SPICE, or a large database. | Let's hope these clericals are doomed, or the PC industry will die as | they flourish. Sorry, you're making poor use of your clerical workers. SPICE they don't need, but access to databases, color highlighting for better productivity, and bit mapped graphics to display pictures and drawings all boost productivity. -- bill davidsen (davidsen@crdos1.crd.GE.COM -or- uunet!crdgw1!crdos1!davidsen) "The world is filled with fools. They blindly follow their so-called 'reason' in the face of the church and common sense. Any fool can see that the world is flat!" - anon
davidsen@crdos1.crd.ge.COM (Wm E Davidsen Jr) (10/23/89)
In article <318@ssp1.idca.tds.philips.nl>, roelof@idca.tds.PHILIPS.nl (R. Vuurboom) writes: | Racing cars today | no longer represent the forefront of passenger car technology but have | become a separate technology. I admit that I have taken this out of context, but what are you driving? The cars on your showroom floor are loaded with stuff which was racing only 20 years ago, like fuel injection, supercharging (and turbocharging), 4 valves/cylinder, overhead cams, etc. These things are not on the sports models only, but in the sedans and station wagons. I think that micros will continue to enbody features developed in more expensive systems. Note the trend to Harvard archetecture, pipelines, fewer cycles per instruction, etc. New designs are adding the f.p. to the main CPU, while keeping it as a coprocessor in the sense of operation in parallel with the main CPU. By the turn of the century I expect to see vector processing on the top machines, using the burst memory access modes already in limited use. I don't see the feature migration or more mips ending as trends any time in the next two decades. Advances in graphics need more address space and CPU. All this means more bandwidth needed. Better graphics and more CPU allows more engineering tasks, which in turn needs more CPU. Multitasking o/s's for micros are gaining market share and drive the need for power, as does the availability of large cheap mass storage which leads to bigger disk servers. If AI ever really gets going (someone finds a mass application) that will keep pushing the need for more CPU. There certainly are limitations to demand, we're just not anywhere near them. When displays reach the limit of the human eye to extract information, the growth in that area will stop. The limit is somewhere around 1200 dpi, 24 bits or color, and whatever screen size you need for your working distance. The CPU limit will be reached when any problem can be solved and the answer displayed in less than 20 ms (the time the eye takes to notice a change). I haven't noticed that we're getting particularly close to any of these limits, so I don't expact the demand for progress to stop any time soon. The limiting factor will be the ability of vendors to provide advances at effective costs. -- bill davidsen (davidsen@crdos1.crd.GE.COM -or- uunet!crdgw1!crdos1!davidsen) "The world is filled with fools. They blindly follow their so-called 'reason' in the face of the church and common sense. Any fool can see that the world is flat!" - anon
bzs@world.std.com (Barry Shein) (10/24/89)
> >When industry does show up with a supposed donation either they have > >ulterior motives (let's donate one of our XY1000's so they'll buy more > >of them!) or they're looking to get some cheap programming done under > >the guise of "research" (with a tax deduction to boot!) > >First you say that industry would get what they need from education if they >donated funds and then you say that getting what they need is an ulterior >motive. Were you just venting or did you have a point? > >mike ching Yes, the mere donation of something (anything) does not help the situation. This is particularly true when the donation is in the form of a few computers or a contract to do some specific programming the industrial concern happens to need right now. It might help, but that's usually a coincidence. If they want educational programs they have to consider donating funds to educational programs (eg. funds for staff/faculty salaries, unfettered funds for labs, or at least sit down and find out what will really help the educational programs rather than their own short-term marketing whims.) Companies are always running over to universities to dump whatever it is they're trying to create some noise about today. Sometimes this is helpful but it usually leaves the universities back where they started in regards to things like salaries for qualified staff and faculty and creating meaningful educational programs. In many cases it actually leaves the departments involved in a WORSE situation since no one (politically) can say no to a whiz-bang sounding donation but its arrival only strains staff and faculty more to adapt to the new equipment (no joke, very serious problem, it can be like giving a drunk a free bottle of whiskey, if you'll pardon the analogy.) Tails wagging dogs is the problem. There's a hell of a lot more to creating good educational programs than dumping a few PC's on a table and plugging them in. And companies are usually loathe to get much involved in anything much beyond making some equipment available. Yet they'll complain loudly when they discover universities aren't teaching skills they need. Something's got to give. Now do you get my point? Note: Of course there are admirable exceptions to the above and all that, but I think it's safe to say if the corporations are still complaining then the point must be valid. I don't mean to express any ungratefulness to those companies that do understand and do try to help. -- -Barry Shein Software Tool & Die, Purveyors to the Trade | bzs@world.std.com 1330 Beacon St, Brookline, MA 02146, (617) 739-0202 | {xylogics,uunet}world!bzs
roelof@idca.tds.PHILIPS.nl (R. Vuurboom) (10/24/89)
In article <1322@crdos1.crd.ge.COM> davidsen@crdos1.UUCP (bill davidsen) writes: >In article <318@ssp1.idca.tds.philips.nl>, roelof@idca.tds.PHILIPS.nl (R. Vuurboom) writes: >| Racing cars today >| no longer represent the forefront of passenger car technology but have >| become a separate technology. > > I admit that I have taken this out of context, but what are you >driving? The cars on your showroom floor are loaded with stuff which was >racing only 20 years ago, like fuel injection, supercharging (and >turbocharging), 4 valves/cylinder, overhead cams, etc. These things are >not on the sports models only, but in the sedans and station wagons. > But isn't this because racing car technology 20 years ago did indeed represent the forefront of passenger car technology. This is exactly my point. This _domination_ of performance boosting (and the resulting performance boosting techniques) has ended in the passenger car technology of today. Which doesn't mean to say that a lot of work isn't going on in this area. It is exactly because perfomance boosting was one of the dominating attributes that racing car technology represented a forefront of passenger car technology. > I think that micros will continue to enbody features developed in more >expensive systems. In the sense that this is a general engineering trend I also think that this will always be the case. > > I don't see the feature migration or more mips ending as trends any >time in the next two decades. Neither do I. Performance boosting in the automobile industry was a dominating factor for some 70 years. I don't see any particular reason why the computer industry will do it in less in fact assuming the same maturity curve we would still have around 30 years to go before this attribute is no longer dominant. > There certainly are limitations to demand, we're just not anywhere >near them. When displays reach the limit of the human eye to extract >information, the growth in that area will stop. The limit is somewhere >around 1200 dpi, 24 bits or color, and whatever screen size you need for >your working distance. The CPU limit will be reached when any problem >can be solved and the answer displayed in less than 20 ms (the time the >eye takes to notice a change). I haven't noticed that we're getting Agreed with the minor point that I consider the human reaction time (200 ms) to be the determining factor. But why quibble over a mere factor of 10? :-) -- Artificial Intelligence: When computers start selling stocks because its Friday the 13th.... Roelof Vuurboom SSP/V3 Philips TDS Apeldoorn, The Netherlands +31 55 432226 domain: roelof@idca.tds.philips.nl uucp: ...!mcvax!philapd!roelof
davidsen@crdos1.crd.ge.COM (Wm E Davidsen Jr) (10/24/89)
In article <6626@ficc.uu.net>, peter@ficc.uu.net (Peter da Silva) writes: | A large part of the problem is that the operating system on this 8088-based | PC was designed to allow one person to use the computer at a time. A network | is a multi-user environment, and without a multi-user computer (multitasking | with at least application-level protection) you can't make effective use of | one. Actually the old XT will run UNIX nicely, being quite a bit bigger than the original PDP-11. We still have XT's and AT's running PC/IX (SysIII) because they do what is needed and no one can cost justify an upgrade. The hardware protect is adequate for a non-hostile environment, I have yet to see anyone reset the segment registers by accident, and for a system to allow a few people to read mail and news they were fine. Not that I'd ever go back, but the hardware is capable of more than MS-DOS will ever get out of it. -- bill davidsen (davidsen@crdos1.crd.GE.COM -or- uunet!crdgw1!crdos1!davidsen) "The world is filled with fools. They blindly follow their so-called 'reason' in the face of the church and common sense. Any fool can see that the world is flat!" - anon
peter@ficc.uu.net (Peter da Silva) (10/25/89)
In article <1419@crdos1.crd.ge.COM> davidsen@crdos1.UUCP (bill davidsen) writes: > Actually the old XT will run UNIX nicely, being quite a bit bigger > than the original PDP-11. I know. I've been there. > The hardware protect is adequate for a non-hostile environment, And it's adequate for a mildly-hostile (need a better term... one where the users are freindly but the software is clumsy or hostile) environment where you can't create executable files at will. Application level protection, in short. -- Peter da Silva, *NIX support guy @ Ferranti International Controls Corporation. Biz: peter@ficc.uu.net, +1 713 274 5180. Fun: peter@sugar.hackercorp.com. `-_-' "That particular mistake will not be repeated. There are plenty of 'U` mistakes left that have not yet been used." -- Andy Tanenbaum (ast@cs.vu.nl)
gillies@m.cs.uiuc.edu (10/25/89)
I am sick and tired of hearing comparisons with the auto industry. CARS PCs ---------------------------------------------------------------------------- Lifespan prop. to miles prop. to age, independent of on-time Failure almost total failure almost never fails totally Mode at 100-140K miles wears very unevenly; most parts (most types of cars) have longer life expectancy than wears evenly a car's. Wear points engine/body/brakes/ Hard Disk/Keyboard/Power Supply tires/interior Cost of fuel increasing constant Source of Mostly Custom Mostly standard/commodity parts Parts (Hardware) Custom (Software)...but software doesn't wear out! Repair off-market 5 yrs off-market X yrs parts after last model after last model from company X from ALL companies, but plug-compatible replacement parts may still be available later. Reason to age; safety Envy of another system's software/speed junk increases; risk of AND no upgrade path (3rd-party vendors driving somewhere often provide upgrades for popular without spare parts "ophan" machiens (XT;Mac128K;Lisa))
mslater@cup.portal.com (Michael Z Slater) (10/26/89)
I don't think the analogy to cars holds up at all. There is little or no utility in continuing to increase the speed of cars, since they are limited by safe driving conditions and roads. The analogy is fundamentally flawed. I think that there is a tremendous need for more performance, and that it will benefit users at all levels. If you look back in time a bit, once all computer displays were character-oriented, since there just wasn't enough compute power to effectively handle a bit-mapped display. Early word processors didn't handle fonts, for example, because it too too much compute power to provide reasonable performance (display technology is also an issue here, but nevertheless the need for more compute power was evident). It seems naive at best to assume that we have reached the pinnacle of user interface design and interactive technology, and it is almost axiomatic taht more sophisticated interfaces will require more compute power. Note that "more sophisticated" may mean _easier_ to use. The lack of a need for fast CPUs for clerical workers is true only if you take today's software as a given. Future software will make all levels of workers more productive, and will require faster CPUs. The availablility of faster CPUs at an affordable price will eventually result in a qualitative change in the way applications work, and in how productive users can be. Michael Slater, Microprocessor Report mslater@cup.portal.com
jhm+@andrew.cmu.edu (Jim Morris) (10/31/89)
Excellent post! I would only add that FAX handles image, but maybe you would include that in the "easy to use" part.
ajs33293@uxa.cso.uiuc.edu (11/10/89)
/* Written 7:37 pm Oct 18, 1989 by gillies@p.cs.uiuc.edu in uxa.cso.uiuc.edu:comp.arch */
/* ---------- "Fed up with MIPS" ---------- */
About being "fed up with MIPS"
Two years ago I bought a Mac II, which was fast as blazes at WYSWYG
word processing. It blew the doors off the old DLION I used to use.
It compiled at 45,000 lines per minute. To friends, I predicted the
hunger for PC cycles would drop in the next five years. Most people
wouldn't need more than 3 MIPs.
But today, a 1000*1000*24bit graphics display takes at least 10 MIPS
to refresh interactively. Today's Mac II struggles at 3-4 MIPS and
1/4 the resolution. It is now "the premier platform for photographic
visual arts."
But most office clerical work needs no more than an XT 8088 machine
and WordPerfect. They don't need color, SPICE, or a large database.
Let's hope these clericals are doomed, or the PC industry will die as
they flourish.
To keep the PC industry thriving we must produce many more technicians
and design engineers, to use high-MIPS CPUs in design optimization and
simulation. There is a big possibility that the U.S. educational
system will let us down. Science and Engineering are unpopular
subjects; everyone wants to be a business major.
So I'm again predicting doom for the PC industry, but this time for
different reasons.
Don Gillies, Dept. of Computer Science, University of Illinois
1304 W. Springfield, Urbana, Ill 61801
ARPA: gillies@cs.uiuc.edu UUCP: {uunet,harvard}!uiucdcs!gillies
/* End of text from uxa.cso.uiuc.edu:comp.arch */asg@pyuxf.UUCP (alan geller) (11/13/89)
In article <112400004@uxa.cso.uiuc.edu>, ajs33293@uxa.cso.uiuc.edu writes: > > /* Written 7:37 pm Oct 18, 1989 by gillies@p.cs.uiuc.edu in uxa.cso.uiuc.edu:comp.arch */ > /* ---------- "Fed up with MIPS" ---------- */ > About being "fed up with MIPS" > > ... discussion about how engineering users need powerful platforms, but > ... clericla users only need PC-XT-level power. > > To keep the PC industry thriving we must produce many more technicians > and design engineers, to use high-MIPS CPUs in design optimization and > simulation. There is a big possibility that the U.S. educational > system will let us down. Science and Engineering are unpopular > subjects; everyone wants to be a business major. > > So I'm again predicting doom for the PC industry, but this time for > different reasons. > > Don Gillies, Dept. of Computer Science, University of Illinois Everyone wanting to be a business major isn't so bad. True, they won't want to run SPICE, or fancy CAD software, but they WILL want to run large spreadsheets, and they'll want results in seconds, not minutes, and they'll want color 3-D graphs, not tables of raw numbers. Don mentions the Mac II; he should try running Wingz on anything less. Even on a Mac II, using Wingz makes you want a IIx or IIcx, and from what I can see, Wingz and 'presentation' spreadsheets like it are the wave of the future, for MBAs anyway. Actually, I've found that management types also like programs such as PowerPoint, which requires real horsepower; FullWrite or Word 4.0 instead of a simpler, smaller word processor; and, coming soon, decision support systems on desktop micros. I hope!! Alan Geller Regal Data Systems ...!{princeton,rutgers}!bellcore!pyuxf!asg