marknew@rosevax.Rosemount.COM (Mark Newman) (03/06/90)
Last week I posted a message asking what was wrong with the '286. A
number of people were kind enough to respond with some useful infor-
mation and opinions, and several asked that I post a summary. Well,
here it is. I edited some of the responses to keep the length of the
summary reasonable.
-Mark Newman
marknew@rosevax.rosemount.com
========================================================================
From Raymond Chen (raymond@math.berkeley.edu):
There is no way to switch from protceted mode to real mode on a 286.
The way it's done in practise is [1] save away the information you
want to preserve, [2] set a flag to say "I am doing something goofy",
[3] tell the keyboard chip to send a reset signal, [4] then halt the
processor. Then when the reset signal comes (after a few milli-
seconds), [5] the chip restarts in real mode, [6] notices the "I am
doing something goofy" flag, [7] loads the context from the stuff
you saved away in step [1] and continues.
Gordon Letwin has called it "Stopping the car in order to change
gears."
- - -
From Dean William Wallwey (wallwey@snoopy.colorado.edu):
You definitely want to get a 386SX. I have a 286, and I am very
soon going to upgrade my mother board to a 386SX because 286's are
going to be left behind. There are a lot of reasons to buy the 386SX.
The 386 and 386SX both have very powerful memory mapping powers which
are used by the new operating systems. They also both have what INTEL
calls Virtual 86 mode so you can multitask DOS programs using Windows
or DesqView for the 386. The small amount of money you save by buying
a 286 machine will come back to haunt you tommorrow, even if you use
it at home. The new software that is coming around the bend really
needs large flat memory spaces (EVEN the NEW 386 version of OS/2
later this year will take advantage of the 386's powers and really
shine!)
Some background info:
The 386SX is identical to the 386 from a software point of view.
The 386SX uses a 16 bit bus whereas the 386 uses a 32 bus so the
386 is a little faster at the same speed because the 386SX has to
reference memory twice for a 386's every one. Both the 386 and 386SX
microprocessors are 32 bits internally and act exaxctly the same. The
286 is a 16 bit bus and a 16 bit architecture. It does not have the
powerful memory managment facilities of the 386. The 286 is *BRAIN
DEAD* compared with 386 and 486 computers. The 386,386SX and 486 will
run OS/2, mulitasking DOS, and UNIX, MUCH MUCH better than the 286
ever will. Even the new Windowing word processors from MicroSoft
(WORD for WINDOWS already out) and the future WordPerfect for OS/2
will run so much better on a 386/386SX than the 286.
Even if you don't think that you'll need the power of the 386 group,
you will soon see even the simple "home" programs like word pro-
cessing and spreadsheets longing for the 386/386SX. (I guess they
already do with WORD for WINDOWS, and Lotus 1-2-3 ver 3.0 even though
they will "run" on a 286).
- - -
From Michael D. Kersenbrock (michaelk@copper.WR.TEK.COM):
The 286 is brain dead primarily because Intel made two major errors
in it's design:
1.) In protected mode (where it can access more than 1 Megabyte),
the chip's MMU-ish architecture is "different" to the point
that DOS can't handle it. DOS MUST run in "real" mode (1Meg
addressablity only). When running DOS, protected mode IS
entered sometimes, but interrupts have to be turned off during
the time it is in protected mode.
2.) Given #1 above: when protected mode is entered, Intel provides
NO way to go BACK to real mode. Because real mode is the mode
that the chip powers up in, what an AT does to return into
REAL mode is to have itself hardware-reset (by the keyboard
controller I think) then, "recover" from where it left off.
Pretty wild.
So, it's not brain dead in the sense of being useless, but brain dead
in the sense that it isn't what it could/should have been had things
been done just a little bit different.
BTW - I just traded some stuff for a used 12-Mhz AT backplane, I hope
to get it running this week (I've a Turbo-XT clone). Even with a
brain dead processor, it should run two or three times faster than my
old PC....
BTW#2- Actually my old 10Mhz XT clone is adequate for home use (given
also that it has 3Meg of EMS 4.0 memory that helps it a LOT).
- - -
From Marc Louis Levinson (gt0159a@prism.gatech.edu):
The 286 is not really brain dead. The problem is that nobody has
ever properly taken advantage of its capabilities. The chip has
available what is called protected mode, which allows advanced
memory handling and multitasking. Chances are slim that you can ever
find a useful DOS application to use this mode.
OS/2, UNIX and some Network software use 286 protected mode. The 386
and 386sx can use advanced memory handling for normal DOS. Overall,
the 386 series has faster throughput.
If you are not planning to use hi power applications, like LOTUS 3,
Windows [loaded to the max], Autocad and others to maximum perform-
ance abilities, then a 286 is ok. Realize that much of the software
now being developed will not run on a 286 but will run on an sx. Con-
sider the lifetime of your computer before you buy one.
- - -
From Dick O'Connor (djo7613@blake.u.washington.edu):
There's nothing wrong with the '286. You should see the programs my
wife and I did for the kids, for household organization, for
gardening, for the church, on and on with an 8 MHz XT-compatible, and
our new 286 machine screams thru all of them. No way we need a 386
for home use, and few people at my office need them either.
I can't speak for all of the mags, who have a vested interest in
pleasing Intel, but I personally feel Bill Machrone and the editorial
staff of PC Magazine are for the most part the most arrogant bunch of
trend-setter-wanna-be's I've ever had the agony of reading. They care
not for the actual uses of the individual, but only for unlimited
desktop power. And those dudes use their machines for word pro-
cessing! What a laugh...
Well, I shouldn't lump them all in the same boat. Machrone is
arrogant, but Seymour is right about half the time, Dvorak *knows*
he's trying to needle you, so I can take him, Manes is funny, and
Zachmann has just guessed wrong in a BIG way about OS/2. Then again,
they could all be the prophets of a new age, and I could be dead
wrong!
- - -
From John Limpert (johnl@n3dmc.UU.NET):
The 80286 cannot deal with numbers larger than 16 bits or data larger
than 64K bytes without doing it in ugly, slow and error prone soft-
software. The 80386 can work with 32 bit numbers and 4 gigabyte data
areas (with the right operating system and software). Many programs
were designed for 32 bit machine like the 80386 and if they will run
at all on the 80286 (and many wont), they operate very slowly. Don't
buy an 80286 machine, you will be wasting your money. In the future
there will be a lot of software that will only be available for 80386
class machines.
I am writing this on the 80286 UNIX system I bought several years
ago. I wouldn't throw it out but I wouldn't buy another one today.phil@pepsi.amd.com (Phil Ngai) (03/08/90)
In article <8681@rosevax.Rosemount.COM> marknew@rosevax.UUCP (Mark Newman) writes: |From Raymond Chen (raymond@math.berkeley.edu): | There is no way to switch from protceted mode to real mode on a 286. Of course not. What good is a protected mode if a random program can violate it? |From Dean William Wallwey (wallwey@snoopy.colorado.edu): | The 386 and 386SX both have very powerful memory mapping powers which | are used by the new operating systems. What new operating systems are you thinking of? Unix, perhaps? | calls Virtual 86 mode so you can multitask DOS programs using Windows | or DesqView for the 386. The small amount of money you save by buying I still don't believe the DV386 uses virtual 86 mode. Does anyone have any definitive data? Maybe I'll ask QD. | The new software that is coming around the bend really | needs large flat memory spaces (EVEN the NEW 386 version of OS/2 | later this year will take advantage of the 386's powers and really | shine!) It doesn't seem to me that there is very much software for OS/2. I know of Excel and Pagemaker. That's about it. Windows, on the other hand, has a lot of software. And when 3.0 comes out, it will run in protected mode, allowing 16 megabytes of memory, ON A 286. | The 386,386SX and 486 will | run OS/2, mulitasking DOS, and UNIX, MUCH MUCH better than the 286 | ever will. Even the new Windowing word processors from MicroSoft Unix does run better on 386 and 486. I don't know what you mean by multi-tasking DOS. | (WORD for WINDOWS already out) and the future WordPerfect for OS/2 | will run so much better on a 386/386SX than the 286. This is untrue. Word for Windows does not take advantage of the 386. Windows 3.0 will take advantage of the 286. | Even if you don't think that you'll need the power of the 386 group, | you will soon see even the simple "home" programs like word pro- | cessing and spreadsheets longing for the 386/386SX. (I guess they | already do with WORD for WINDOWS, and Lotus 1-2-3 ver 3.0 even though | they will "run" on a 286). I don't think you know what you are talking about. |From Michael D. Kersenbrock (michaelk@copper.WR.TEK.COM): | The 286 is brain dead primarily because Intel made two major errors | in it's design: | 1.) In protected mode (where it can access more than 1 Megabyte), | the chip's MMU-ish architecture is "different" to the point | that DOS can't handle it. DOS MUST run in "real" mode (1Meg This seems like a problem with DOS, since the architecture of the 8086 is inherently limited to 1 megabyte, any improvements would have to be different, and it's up to DOS to take advantage of it. In addition, MS has said there will be a DOS 5 and DOS 6 (in InforWorld). Perhaps by then they will get around to using the power of the 286. | So, it's not brain dead in the sense of being useless, but brain dead | in the sense that it isn't what it could/should have been had things | been done just a little bit different. Nonsense. Tell me specifically what you would have done differently from Intel? I bet you can't come up with anything. |From Marc Louis Levinson (gt0159a@prism.gatech.edu): | The 286 is not really brain dead. The problem is that nobody has | ever properly taken advantage of its capabilities. The chip has | available what is called protected mode, which allows advanced | memory handling and multitasking. Chances are slim that you can ever | find a useful DOS application to use this mode. At last, someone who knows what he's talking about. | If you are not planning to use hi power applications, like LOTUS 3, | Windows [loaded to the max], Autocad and others to maximum perform- | ance abilities, then a 286 is ok. Realize that much of the software | now being developed will not run on a 286 but will run on an sx. Con- | sider the lifetime of your computer before you buy one. With Windows 3.0, which does run in protected mode on a 286, due to come out early this summer, you should consider exactly what the extra cost of a 386 will buy you. Also realize that 10 or 20 million 286 machines constitutes a powerful reason for software companies to support the 286. |From John Limpert (johnl@n3dmc.UU.NET): | Many programs | were designed for 32 bit machine like the 80386 and if they will run Name some, besides CAD packages. I think most people want things like spreadsheets, word processors, and games. These will run just fine on a 286. | In the future | there will be a lot of software that will only be available for 80386 | class machines. Again, there is a very large market for 286 applications and they simply will not be left unsupported. | I am writing this on the 80286 UNIX system I bought several years | ago. I wouldn't throw it out but I wouldn't buy another one today. Certainly Unix is one program that benefits from a 386. But users don't buy PCs to run Unix, they buy PCs to run applications, and Unix is one of the worst systems to try to buy applications for. What kind of spreadsheets can you get? Anything that comes even close to Excel? I hear you can finally get MS Word 5.0 on Unix. A character based application, when DOS has Word for Windows. If you've got the Unix religion, I guess you'll get a 386. But if you want to get work done, you'll stay away from Unix. -- Phil Ngai, phil@amd.com {uunet,decwrl,ucbvax}!amdcad!phil A PC without DESQview is like Unix without ^Z.
jmerrill@jarthur.Claremont.EDU (Confusion Reigns) (03/08/90)
In article <29405@amdcad.AMD.COM> phil@pepsi.AMD.COM (Phil Ngai) writes: >In article <8681@rosevax.Rosemount.COM> marknew@rosevax.UUCP (Mark Newman) writes: >|From Raymond Chen (raymond@math.berkeley.edu): >| There is no way to switch from protceted mode to real mode on a 286. > >Of course not. What good is a protected mode if a random program >can violate it? The idea of a protected mode as I see it is not to keep programs from doing anything -- what would be the point? It was intended to keep 8086 programs from getting out of line. The 286's lack of a protected -> real switch is not an obstacle to any 'random program'; observe the hack involving resetting the chip. Your 'random program' could do that just as easily as send the protected -> real instruction to the 386. You gain nothing by not implementing a switch from protected to real modes. >| (WORD for WINDOWS already out) and the future WordPerfect for OS/2 >| will run so much better on a 386/386SX than the 286. > >This is untrue. Word for Windows does not take advantage of the 386. >Windows 3.0 will take advantage of the 286. I think the issue is speed. The 386 has a 32-bit bus; the 286 has a 16-bit. For memory-intensive programs, this is important. >| So, it's not brain dead in the sense of being useless, but brain dead >| in the sense that it isn't what it could/should have been had things >| been done just a little bit different. > >Nonsense. Tell me specifically what you would have done differently >from Intel? I bet you can't come up with anything. I would have made it a 386. Or a 486. Or... >|From Marc Louis Levinson (gt0159a@prism.gatech.edu): >| The 286 is not really brain dead. The problem is that nobody has >| ever properly taken advantage of its capabilities. The chip has >| available what is called protected mode, which allows advanced >| memory handling and multitasking. Chances are slim that you can ever >| find a useful DOS application to use this mode. > >At last, someone who knows what he's talking about. So only people who share your opinions know what they are talking about? Gee...how are you on politics? Religion? >A PC without DESQview is like Unix without ^Z. And DESQview without a 386 is like Unix without bg. How many of us have EEMS boards? -- Jason Merrill jmerrill@jarthur.claremont.edu Disclaimer: Feel free to ignore everything I say.
ralf@b.gp.cs.cmu.edu (Ralf Brown) (03/08/90)
In article <4983@jarthur.Claremont.EDU> jmerrill@jarthur.Claremont.EDU (Confusion Reigns) writes: }from getting out of line. The 286's lack of a protected -> real switch is }not an obstacle to any 'random program'; observe the hack involving }resetting the chip. Your 'random program' could do that just as easily as }send the protected -> real instruction to the 386. Not quite the same. The 386 protected->real switch can only be executed by a process running at the greatest (kernel) privilege. And on a 386, the I/O space can be virtualized such that unprivileged processes can't tell the keyboard controller to reset the CPU. -- {backbone}!cs.cmu.edu!ralf ARPA: RALF@CS.CMU.EDU FIDO: Ralf Brown 1:129/46 BITnet: RALF%CS.CMU.EDU@CMUCCVMA AT&Tnet: (412)268-3053 (school) FAX: ask DISCLAIMER? | _How_to_Prove_It_ by Dana Angluin 3. by vigorous handwaving: What's that?| Works well in a classroom or seminar setting.
marshall@wind55.seri.gov (Marshall L. Buhl) (03/09/90)
>Phil Ngai "debunks" anti-286 sentiment in a previous post.
Phil,
Once again we find ourselves on differing sides of an issue. The 286
is costing users and developers a fortune. Trying to shoehorn modern
applications into a 16-bit address space is very expensive. This drives
up software costs for users. Developers have to develop for two platforms.
One for people who have to get work done and one for people who don't
value their time.
Sure, you can save a couple hundred on a 286. But what is that compared
to the cost of the ENTIRE system - hardware AND software. Peanuts.
I run Windows on a 386/25. It's a DOG! I was a beta tester for WinWord.
I own it, but don't use it. My time is just too valuable to have to wait
for graphics screen refreshes all the time. I've gone back to using PC
Word, because I don't have a PC powerful enough to run WinWord. I just
can't imagine how miserable this fantastic product would be on a 286.
Phil. Please throw that old junker on the trash heap. People like you
are holding the rest of us back. At least send it home with someone
as I do with my old junkers. Get it off you desk. There is no place
in corporate America for the 286.
Marshall
PS. They DO make good consoles for multi-user *nix boxes. I put them
in screen capture mode so I have a record of all console messages.
Be sure to save that file periodically and start a new file. If
it gets too big, that 286 will take forever to search for strings
when you want to review the messages.
PPS. The 286 is dead. Short live the 386 and 486. Bring on the 586.
I need it today!!! 8-)
--
Marshall L. Buhl, Jr. EMAIL: marshall@wind55.seri.gov
Senior Computer Engineer VOICE: (303)231-1014
Wind Research Branch 1617 Cole Blvd., Golden, CO 80401-3393
Solar Energy Research Institute Solar - safe energy for a healthy futuremarshall@wind55.seri.gov (Marshall L. Buhl) (03/09/90)
jmerrill@jarthur.Claremont.EDU (Confusion Reigns) writes: >So only people who share your opinions know what they are talking about? >Gee...how are you on politics? Religion? DON'T ask him about guns either! -- Marshall L. Buhl, Jr. EMAIL: marshall@wind55.seri.gov Senior Computer Engineer VOICE: (303)231-1014 Wind Research Branch 1617 Cole Blvd., Golden, CO 80401-3393 Solar Energy Research Institute Solar - safe energy for a healthy future
wallwey@boulder.Colorado.EDU (WALLWEY DEAN WILLIAM) (03/09/90)
In article <29405@amdcad.AMD.COM> phil@pepsi.AMD.COM (Phil Ngai) writes: >In article <8681@rosevax.Rosemount.COM> marknew@rosevax.UUCP (Mark Newman) writes: >|From Raymond Chen (raymond@math.berkeley.edu): >| There is no way to switch from protceted mode to real mode on a 286. > >Of course not. What good is a protected mode if a random program >can violate it? > The way OS/2 switches back to real mode on a 286 is to have the keyboard controller actually RESET the 286! The good of it, is so you can still run you DOS programs, like Lotus and WordPerfect!!!!! >|From Dean William Wallwey (wallwey@snoopy.colorado.edu): >| The 386 and 386SX both have very powerful memory mapping powers which >| are used by the new operating systems. > >What new operating systems are you thinking of? Unix, perhaps? OS/2 ver 2.0 which will require a 386, Windows386 ver 3.0 and of course the "real" UNIXs you can run on a PC. > >| calls Virtual 86 mode so you can multitask DOS programs using Windows >| or DesqView for the 386. The small amount of money you save by buying > >I still don't believe the DV386 uses virtual 86 mode. Does anyone have >any definitive data? Maybe I'll ask QD. > I'm pretty sure DV386 uses the virtual 86 mode of the 386, that's why you need a 386 to run DV386. Likewise with Windows 386. This is the only way you can reliably multitask regular DOS programs!!!! Most DOS programs (like WP's and SpreadSheets) think and need complete control of your system so they write directly to your screen and lots of other "nasty" things. Vitual mode lets the programs think they are still able to do this, but actually don't! By the way OS/2 ver 2.0 is also supposed to support this feature so OS/2 Ver 2.0 can multitask OS/2 ver 1.0 programs, OS/2 ver 2.0 programs, DOS programs, and if the rummors are right Windows 3.0 programs all at the same time and on the same screen. But you will need a 386 to run OS/2 ver 2.0!!!!!!!!! >| The new software that is coming around the bend really >| needs large flat memory spaces (EVEN the NEW 386 version of OS/2 >| later this year will take advantage of the 386's powers and really >| shine!) > >It doesn't seem to me that there is very much software for OS/2. >I know of Excel and Pagemaker. That's about it. You should read some more PC magazines. Not only is Microsoft strongly commited to OS/2, but so are WordPerfect Corp, as well as Lotus, and AutoCad coming out with versions of there most popular software for OS/2 that will make their DOS versions look like they came from the Dark Ages!!!!!! Even Excel, is so much better for OS/2 then Winodws!!! > >Windows, on the other hand, has a lot of software. And when 3.0 >comes out, it will run in protected mode, allowing 16 megabytes of >memory, ON A 286. Granted, but most of the Wonderful software from Lotus, WP, Autocad, and yes even Microsoft will run much faster, cleaner, and just plan better on a 386. Even Windows 386 ver 3.0 will be MUCH BETTER than Windows 286 ver 3.0!!!!!!! By the way most software companies that write applications for Windows are in the process of rewriting their code for OS/2, Especially MicroSoft!!! > >| The 386,386SX and 486 will >| run OS/2, mulitasking DOS, and UNIX, MUCH MUCH better than the 286 >| ever will. Even the new Windowing word processors from MicroSoft > >Unix does run better on 386 and 486. I don't know what you mean by >multi-tasking DOS. See above----Try running a BBS, Word Perfect, and Lotus, and MicroSoft Windows on a 286 at the same time, and then try it on a 386! I would be suprized if you could do it on a 286, and if even if you could, if any one program locked up, it would take your whole system with it. Not on a 386 running DV386, Windows386 or the soon to be OS/2 ver 2.0 (386)!!!! Also Windows 386 offers TRUE multitasking in the sense of Time Slicing, Windows 286, uses co-operative multitasking. > >| (WORD for WINDOWS already out) and the future WordPerfect for OS/2 >| will run so much better on a 386/386SX than the 286. > >This is untrue. Word for Windows does not take advantage of the 386. >Windows 3.0 will take advantage of the 286. Note I said will "run better on a 386" rather than take advantage of of the 386. Windows 386, as will Windows 386 ver 3.0, is still much faster because of the special memory mapping modes of the 386. Granted Windows 286 ver 3.0 will be able to use the protected mode memory of the 286 but it can't map it as powerfully as the 386, and hence will be slower, not to mention 386 machines are generally faster to begin with. > >| Even if you don't think that you'll need the power of the 386 group, >| you will soon see even the simple "home" programs like word pro- >| cessing and spreadsheets longing for the 386/386SX. (I guess they >| already do with WORD for WINDOWS, and Lotus 1-2-3 ver 3.0 even though >| they will "run" on a 286). > >I don't think you know what you are talking about. I don't think you know what you are talking about! If you read many of the news groups, people are complaining seriously about how slow Word for Windows is on a 286. Many people consider it to slow to actually use on a 286. Also what I ment by my prevous comment is, that programs are getting bigger, and actually requiring more speed due to programs's complexity, and the higher resolution display's and graphics environments. The 386 is "better" suited for these. Finally I have worked on 386 machines, and own a 286. I also do MsWindows Software Development. I hate my 286!!!!! and I am going to move up to a 386 as soon as possible. I DON'T WANT TO SEE ANYBODY ELSE MAKE THE MISTAKE OF BUYING A 286 WHEN THEY COULD BUY A 386 FOR $100-$200 MORE!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! Also if you don't think my opinions are worth anything, check out some of Charles Petzolds comments in PC Magazine about how he feels about the 286 vs 386 issue. He has written a couple of OS/2 and MSWindows books for MicroSoft and is considered a great PM and Windows programer by most developers! He also has a lot of connections at MicroSoft, and if anybody is qualified to comment about future software delevelopments and corporate strategy, he is the one. > >|From Michael D. Kersenbrock (michaelk@copper.WR.TEK.COM): >| The 286 is brain dead primarily because Intel made two major errors >| in it's design: >| 1.) In protected mode (where it can access more than 1 Megabyte), >| the chip's MMU-ish architecture is "different" to the point >| that DOS can't handle it. DOS MUST run in "real" mode (1Meg > >This seems like a problem with DOS, since the architecture of the 8086 >is inherently limited to 1 megabyte, any improvements would have to >be different, and it's up to DOS to take advantage of it. In addition, >MS has said there will be a DOS 5 and DOS 6 (in InforWorld). Perhaps >by then they will get around to using the power of the 286. > Why Wait, even if there is a DOS5 and a DOS6, the 286 can not EVER make regular DOS programs DIRECTLY take advantage protected mode memory. The 386 can do it today with DV386 and Windows 386 not to mention OS/2 ver 2.0 and UNIX !!!!!!!!!!! Granted you can use EMS LIM 4.0 emulators to use the extended protected memory to emulate exapned memory in DOS, but it is so slow, that it is usually faster to user an 8088 with real EMS LIM4.0 memory on memory intensive applications. >| So, it's not brain dead in the sense of being useless, but brain dead >| in the sense that it isn't what it could/should have been had things >| been done just a little bit different. > >Nonsense. Tell me specifically what you would have done differently >from Intel? I bet you can't come up with anything. O' contraire.... All the Things that they should have put into the 286, they actually put on the 386, Like allowing DOS to directly use the extended protected memory by special mapping functions. Having a flat address space, so programs can break the 1982 limit of 64k per data and code segment. When Intel designed the 8086 and carried over many of its bad traits to 286, the designers thought that 64K was alot of memory. Today small windows on the screen can take more than that to just display their associated pixels. Although it can be done, these limits of changing segments all the time, make programs considrably slower than they would be if they used a flat address space! > >|From Marc Louis Levinson (gt0159a@prism.gatech.edu): >| The 286 is not really brain dead. The problem is that nobody has >| ever properly taken advantage of its capabilities. The chip has >| available what is called protected mode, which allows advanced >| memory handling and multitasking. Chances are slim that you can ever >| find a useful DOS application to use this mode. > >At last, someone who knows what he's talking about. > Not really, see my last comments. To use protected mode, the program must actually be written specifically for it. In the case of Windows 3.0 it will use protected mode because it was written for it, and apllications that run under it will have to have been written according to MicroSoft's standards for windows programs-(previous versions of Windows, prior to Ver 3.0 did memory managment by software and used the nearly same memory managment scheme as protected memory. If a windows program conformed to MS's standards, it will be able to use the protected mode available in MsWindows ver 3.0, because we have been writing our programs basically for it all along, and limited ourselves to 64K datasegments.) If you don't beleive me, check out some of Charles Petzolds books and articles again. >| If you are not planning to use hi power applications, like LOTUS 3, >| Windows [loaded to the max], Autocad and others to maximum perform- >| ance abilities, then a 286 is ok. Realize that much of the software >| now being developed will not run on a 286 but will run on an sx. Con- >| sider the lifetime of your computer before you buy one. > >With Windows 3.0, which does run in protected mode on a 286, due to >come out early this summer, you should consider exactly what the >extra cost of a 386 will buy you. A whole lot of headach, I use my 286 with MSWindows, even when Windows 3.0 comes out, it won't be a cure all solution, although I will be very joyed when it finally does come out. Windows 3.0 still won't ba able to truely multitask programs like Word Perfect and Lotus 1-2-3, and a whole lot of other "nasty" DOS programs that have no Windows version counter part! > >Also realize that 10 or 20 million 286 machines constitutes a powerful >reason for software companies to support the 286. > >|From John Limpert (johnl@n3dmc.UU.NET): >| Many programs >| were designed for 32 bit machine like the 80386 and if they will run > >Name some, besides CAD packages. I think most people want things like >spreadsheets, word processors, and games. These will run just fine >on a 286. If you like to run the newest programs super slow!!!!!!!!! > >| In the future >| there will be a lot of software that will only be available for 80386 >| class machines. > >Again, there is a very large market for 286 applications and they >simply will not be left unsupported. Yeh, right, just like the 8088 isn't being left behind today. 386's are cheap and getting cheaper. The Difference between the power of a 286 and a 386 makes up Many times for the small difference in price!! > >| I am writing this on the 80286 UNIX system I bought several years >| ago. I wouldn't throw it out but I wouldn't buy another one today. > >Certainly Unix is one program that benefits from a 386. But users >don't buy PCs to run Unix, they buy PCs to run applications, and >Unix is one of the worst systems to try to buy applications for. >What kind of spreadsheets can you get? Anything that comes even >close to Excel? I hear you can finally get MS Word 5.0 on Unix. >A character based application, when DOS has Word for Windows. If >you've got the Unix religion, I guess you'll get a 386. But if >you want to get work done, you'll stay away from Unix. > I agree UNIX "sucks" for applications. If you want to write real applications, even for OS/2 and Windows, programmers are going to make better versions of their software for the 386!!! Just like Lotus announced 123 ver3.0 for 286's and above, and ver 2.2 for 8088's. You are going to see software targeted, and greatly enhanced for 386 above the 286. >-- >Phil Ngai, phil@amd.com {uunet,decwrl,ucbvax}!amdcad!phil >A PC without DESQview is like Unix without ^Z. Remember the point of my original reply was not to bash the 286, but to help someone not to make a mistake. Also if you think the 286's are so great, you can stay in '80's for all I care, I'm moving into the '90's!!!!!!!!!!!!!!!!!! Dean
sl197009@silver.ucs.indiana.edu (Chima Echeruo) (03/09/90)
I have read many "experts" saying that the 286 is far slower than a 386. I have a 20Mhz 286 and I am yet to find a DOS computer in my college that performs faster than the 286 running DOS. I have read in Byte that there are going to be versions of the 286 at 25 Mhz. Is this true? If it is, won't a 25Mhz 286 outperform an equivalent 386 while in real mode? The only feature that bother me in the 286 is it's lack of "flat memory" and also a lack of 286 specific software. I do not think that the 286 is dead, maybe it is not the state-of-the-art in computer technology but it provides a very *fast* system for the average computer user. If the SX does not go past 16Mhz, I do not see the benefit for upgrading to a 386sx. When the 586 is ready for shipping by Intel, we will hear again how the 386 is brain-damaged and not able to run the lastest sofware that is emerging. When MicroSoft came out with OS/2, PC/XT owners were advised to can their machines and lay down the cash for a 286. After three years not much has been ------------- Chima Echeruo sl197009@silver.ucs.indiana.edu -------------------------------
phil@pepsi.amd.com (Phil Ngai) (03/09/90)
In article <4983@jarthur.Claremont.EDU> jmerrill@jarthur.Claremont.EDU (Confusion Reigns) writes: |The idea of a protected mode as I see it is not to keep programs from doing |anything -- what would be the point? It was intended to keep 8086 programs |from getting out of line. The 286's lack of a protected -> real switch is |not an obstacle to any 'random program'; observe the hack involving |resetting the chip. For a particular box to have a security hole is one thing. But for the processor to have a security hole etched in silicon means that there is no security ever possible. I have had the idea of a switch to real mode in ring 0 suggested, and that sounds kind of interesting. Would anyone use a 286A if it came out? |>| (WORD for WINDOWS already out) and the future WordPerfect for OS/2 |>| will run so much better on a 386/386SX than the 286. |> |>This is untrue. Word for Windows does not take advantage of the 386. |>Windows 3.0 will take advantage of the 286. | |I think the issue is speed. The 386 has a 32-bit bus; the 286 has a |16-bit. For memory-intensive programs, this is important. Uh huh. And what about the 386SX he mentioned as being so superior for running Word for Windows? |>| So, it's not brain dead in the sense of being useless, but brain dead |>| in the sense that it isn't what it could/should have been had things |>| been done just a little bit different. |> |>Nonsense. Tell me specifically what you would have done differently |>from Intel? I bet you can't come up with anything. | |I would have made it a 386. Or a 486. Or... He said a little bit different. |>A PC without DESQview is like Unix without ^Z. | |And DESQview without a 386 is like Unix without bg. How many of us have |EEMS boards? A 386 makes DESQview much easier to run, but so do the All Charge Card, the SOTA Pop, and several other products which are under development. As a matter of fact, a lot of us in my group have AST Premium 286s, with EEMS. But if multi-tasking were so important, I would think QuarterDeck would be almost as big as Microsoft. I don't think they are. A lot of people's reaction is "why do I need multi-tasking?" I don't understand it, but there it is. -- Phil Ngai, phil@amd.com {uunet,decwrl,ucbvax}!amdcad!phil A PC without DESQview is like Unix without ^Z.
wallwey@boulder.Colorado.EDU (WALLWEY DEAN WILLIAM) (03/09/90)
In article <38299@iuvax.cs.indiana.edu> sl197009@silver.ucs.indiana.edu (Chima Echeruo) writes: > >I have read many "experts" saying that the 286 is far slower than a 386. >I have a 20Mhz 286 and I am yet to find a DOS computer in my college that >performs faster than the 286 running DOS. Have you ever tryed a 20MHz 386? It won't be whole lot faster (1.5x), but some. According to some articles I've read, a 16 MHz 286 and a 16Mhz 386SX run very close to the same speed, sometimes with even the 286 being very slightly faster. When you buy a 386SX, you are not buying it for the speed, but for the POWER behind a 386 architechture. If you want speed, go to a true 386 sometimes known as the 386DX. Besides, if you get a 20Mhz 286 you are already paying a lot for the extra 4MHz. > >I have read in Byte that there are going to be versions of the 286 at 25 Mhz. >Is this true? I don't know, but why try to beat a dead horse. > >If it is, won't a 25Mhz 286 outperform an equivalent 386 while in real mode? Not a true 386 (DX). A ture 386, performs many of the same instructions a 286 does in fewer clock cycles, Not to mention, a 387 is MUCH faster than a 287. Also a 386 when in 386 protected mode can hanle true 32 bit quantities, which makes it much faster when using software written for the 386, like AutoCad 386, WINDOWS 386, and soon to be out, OS/2 Ver 2.0, and finally any flavor of UNIX specifically designed for the 386. Sometimes a 286 can beat a 386SX at the same speed, or obviously at a faster rate, but is the extra approx 5% speed difference worth wasting your future computer investment over? >The only feature that bother me in the 286 is it's lack of "flat memory" and >also a lack of 286 specific software. > Also it's lack of Virtual '86 mode allowing true multitasking of real mode DOS applications, and lack of special memory mapping features. >I do not think that the 286 is dead, maybe it is not the state-of-the-art in >computer technology but it provides a very *fast* system for the average >computer user. If the SX does not go past 16Mhz, I do not see the benefit for >upgrading to a 386sx. In one of the latest PC World magazine's I just saw a blurb saying that Intel announced a true 20Mhz 386SX!!!!!!!!!!!! I own a 286 and hate it. I'm going to upgrade my mother board to a 386SX as soon I can. It will probably cost me a total of $300-$400 dollars to do that. Hell, I've spent more than that on a single piece of software a couple of times. Actually in my case I need to upgrade the memory also, and my current motherboard only accepts up to 1MB. To get a memory board alone would cost $200 for a good one. I would rather spend it upgrading my system to a 386SX that can handle the extra memory, and not be limited to the 8Mhz bus speed that the memory expansion board would be. Also in getting a 386SX, I can use any of a number of software packages that can use the 386's ability to map protected memory directly into the DOS area to emulate LIM EMS 4.0, that so many DOS applications require! As for as this thread is concerned, the whole point of this disccussion is whether you should buy a 286 or a 386 if you are buying a new system. Again, the extra $100-$200 dollars spent on a 386SX clone (or better) running at 16MHz around $900-$1100 compared with a 12Mhz 286 clone, around $700-$900 is VERY WELL SPENT, and peanuts compared to costs you will incure while owning a computer system!!!!!!!!!!! As far as the 286 being dead, well its not dead, but why go with 80's tech. when the rest of the world will be going into the 90's for such a small amount more. Also pure speed/$ isn't everything! What runs at that speed IS!!!!!! DOS is from the DARK Ages, OS/2 ver 1.x will be pressed to it's limits even faster than DOS did. > >When the 586 is ready for shipping by Intel, we will hear again how the 386 >is brain-damaged and not able to run the lastest sofware that is emerging. > Agreed, but don't you think the 8086 is brain damaged. When everybody saw the IBM AT with its 286, everybody asked who needed it, 'the 8086 can run all the software out there! Who needs it? ' It all depends on your point of view. This is called progress! >When MicroSoft came out with OS/2, PC/XT owners were advised to can their >machines and lay down the cash for a 286. After three years not much has been >------------- >Chima Echeruo >sl197009@silver.ucs.indiana.edu >------------------------------- Personally that's why I bought my computer, and will for ever buy my computers for a life span of about 3 years. To spend vast amounts on the latest, most expensive equipment is a waste. That's why I'm getting a 386SX. And can't wait till 486's and 586's drop in price. Buy that time I will have saved enough money to get them, and with the 386SX, I can't be limited to software I run (at least for a year or so). The above opinions are clearly my own. They are not ment as a flame. Dean
dhinds@portia.Stanford.EDU (David Hinds) (03/09/90)
In article <38299@iuvax.cs.indiana.edu>, sl197009@silver.ucs.indiana.edu (Chima Echeruo) writes: > > I have read many "experts" saying that the 286 is far slower than a 386. > ... won't a 25Mhz 286 outperform an equivalent 386 while in real mode? From what I've seen, a 16MHz 286 runs DOS code at about the same speed as a 16MHz 386SX, +/- 5% or so. The instruction timings are about the same, and the bus widths are both 16 bits. A 16MHz 386 should be consistently faster, even running 16-bit code and without a RAM cache. The reason is that even if nothing else uses the 32-bit bus bandwidth, the instruction prefetch unit on the 386 always loads whole doublewords. I think the effect of this is something like a 20% improvement in speed over a same-speed 286. The 286 also does not support memory interleaving systems, so for >16MHz, you will be forced to have wait states on main memory. A 20MHz 386 can still work at "near 0 wait states" with interleaved memory. At higher speeds, a static RAM cache is necessary for avoiding wait states. I don't know if there are cache controllers for the 286, but I'm pretty sure the Intel cache controller only works with 386's. > > I do not think that the 286 is dead, maybe it is not the state-of-the-art in > computer technology but it provides a very *fast* system for the average > computer user. If the SX does not go past 16Mhz, I do not see the benefit for > upgrading to a 386sx. > I agree that fast 286's are perfectly good machines for running almost all DOS applications, and are guaranteed to have a high level of support for many years due to the huge number of systems around. As a sort of souped-up 8086, I think it still has a very legitimate market. I think that the 386SX puts a LOT of pressure on new fast 286's (say 16MHz or higher), because the price differential is quite small, both will run DOS code about equally well, but the 386SX offers a bit more. If I had a fast 286 machine, it certainly would not be worth it to me to "upgrade" to a 386SX, but I wouldn't buy another 286. By the way, Intel has already announced the 20MHz 386SX, and a version of their cache controller for the SX. A 20MHz 386SX with memory interleaving or a cache would be significantly faster than a 20MHz 286 with the unavoidable wait states. -David Hinds dhinds@popserver.stanford.edu
jwi@cbnewsj.ATT.COM (Jim Winer @ AT&T, Middletown, NJ) (03/09/90)
> Marshall L. Buhl writes: > Once again we find ourselves on differing sides of an issue. The 286 > is costing users and developers a fortune. Trying to shoehorn modern > applications into a 16-bit address space is very expensive. This drives > up software costs for users. Developers have to develop for two platforms. > One for people who have to get work done and one for people who don't > value their time. Developers don't have to do anything. But it seems to make sense to build programs for the 20 million or so existing machines -- frankly, there aren't enough 386 machines arond to make it a worthwhile market at this time. > I run Windows on a 386/25. It's a DOG! I was a beta tester for WinWord. > I own it, but don't use it. My time is just too valuable to have to wait > for graphics screen refreshes all the time. I've gone back to using PC > Word, because I don't have a PC powerful enough to run WinWord. I just > can't imagine how miserable this fantastic product would be on a 286. I don't understand why you call it a fantastic product when you won't use it because "it's a DOG!" I have done direct comparisons with AMI and WordPerfect and both of them are better products that run just fine on a fast 286. Your problem isn't hardware, it's poor choice of software. > Phil. Please throw that old junker on the trash heap. People like you > are holding the rest of us back. At least send it home with someone > as I do with my old junkers. Get it off you desk. There is no place > in corporate America for the 286. Since most of what corporate America does (word processing, spread sheet, data base, etc.) runs just fine on a 286 with a big enough and fast enough disk, it seems rather dumb to make this statement. > Marshall L. Buhl, Jr. EMAIL: marshall@wind55.seri.gov > Solar Energy Research Institute Solar - safe energy for a healthy future Perhaps if we didn't waste so many things by insisting that they are obsolete when they are not, we wouldn't have as much need for solar energy -- but then that's another area where we can all get gung-ho for hardware for hardware's sake. Jim Winer -- jwi@mtfme.att.com -- Opinions not represent employer. ------------------------------------------------------------------ ...I've had some womderful daydreams about how the FAA controllers would react to suddenly discovering a dragon on short final into O'Hare on a busy night in IFR conditions... -- J.C. Morris, The MITRE Corp., McLean, VA
rob@prism.TMC.COM (03/09/90)
phil@pepsi.UUCP writes: >In article <8681@rosevax.Rosemount.COM> marknew@rosevax.UUCP (Mark Newman) >writes: >|From Raymond Chen (raymond@math.berkeley.edu): >| There is no way to switch from protceted mode to real mode on a 286. >Of course not. What good is a protected mode if a random program >can violate it? This can't be a serious question. Obviously, the means for switching would be via a priviledged instruction, as on the 386/486. The operating system hardly counts as a 'random program'. In fairness, the 286 was designed before anyone knew that DOS compatibility (the main reason for desiring a two-way real <-> protected mode mechanism) would be vital. >|From Dean William Wallwey (wallwey@snoopy.colorado.edu): >| The 386 and 386SX both have very powerful memory mapping powers which >| are used by the new operating systems. >What new operating systems are you thinking of? Unix, perhaps? Also the various DOS multitaskers like Windows 386, VM/386, and so on. There are also numerous 386 DOS extenders. Plus OS/2 version 2, if it ever becomes available (an open question, I'll grant). >| calls Virtual 86 mode so you can multitask DOS programs using Windows >| or DesqView for the 386. The small amount of money you save by buying >I still don't believe the DV386 uses virtual 86 mode. Does anyone have >any definitive data? Maybe I'll ask QD. I'd be interested in knowing for sure, also. The fact that DV386 can virtualize screen access (even running programs that do direct video access) suggests that it _is_ using the V86 mode, but it does seem less stable than I'd expect for a program that uses V86. Still, early versions of Windows 386, which uses V86, were none too stable either. V86 mode makes safe multitasking of DOS applications possible, but not easy. >| The new software that is coming around the bend really >| needs large flat memory spaces (EVEN the NEW 386 version of OS/2 >| later this year will take advantage of the 386's powers and really >| shine!) >It doesn't seem to me that there is very much software for OS/2. >I know of Excel and Pagemaker. That's about it. >Windows, on the other hand, has a lot of software. And when 3.0 >comes out, it will run in protected mode, allowing 16 megabytes of >memory, ON A 286. It will still run better (offering multitasking of DOS applications and better virtual memory support) on a 386. >| The 286 is not really brain dead. The problem is that nobody has >| ever properly taken advantage of its capabilities. The chip has >| available what is called protected mode, which allows advanced >| memory handling and multitasking. Chances are slim that you can ever >| find a useful DOS application to use this mode. >At last, someone who knows what he's talking about. There's a lot of truth to the above paragraph; the 286 has seen use mostly as a fast 8088, which hardly does the chip justice. That's not why people disparage it, though. The problem is that 286 protected mode is far less versatile and powerful than 386 protected mode. That's why there's more 386-specific software around than 286-specific software, despite the fact that the 286 has been around years longer. I agree it's unfortunate that the 286 hasn't been used more fully, but at this point, that's not very relevant. >With Windows 3.0, which does run in protected mode on a 286, due to >come out early this summer, you should consider exactly what the >extra cost of a 386 will buy you. Again, Windows 3.0 will support the 386, offering capabilities not available under the 286 version. And the extra cost of a 386, especially if you're willing to buy a non-name brand, is pretty marginal. >Also realize that 10 or 20 million 286 machines constitutes a powerful >reason for software companies to support the 286. Agreed. A similar argument can be made for the 8086/8. >Certainly Unix is one program that benefits from a 386. But users >don't buy PCs to run Unix, they buy PCs to run applications, and >Unix is one of the worst systems to try to buy applications for. >What kind of spreadsheets can you get? Anything that comes even >close to Excel? I hear you can finally get MS Word 5.0 on Unix. >A character based application, when DOS has Word for Windows. If >you've got the Unix religion, I guess you'll get a 386. But if >you want to get work done, you'll stay away from Unix. You're right about Unix, but you can benefit from a 386 even if you're not running Unix, and that degree of benefit will only increase with time. The 286 does offer a great price/performance ratio, and as a fast 8088, which is all many people need, it's tough to beat. And when you realize that it was designed a decade ago, its flaws become more understandable, though no less annoying. Obviously there's no sense buying a 386 or 486 just to have the latest toy, and someone who really feels that a 286 will address their short and long term needs can buy one with a clean conscience. But the benefits of a 386 are real even today, and will be far moreso in the future. Given the narrowing cost difference between 286 and 386 machines, the group of tasks for which the 286 is the CPU of choice is shrinking, and will continue to do so.
phil@pepsi.amd.com (Phil Ngai) (03/10/90)
In article <1990Mar8.182345.11823@seri.gov> marshall@wind55.seri.gov (Marshall L. Buhl) writes: |Once again we find ourselves on differing sides of an issue. The 286 |is costing users and developers a fortune. Trying to shoehorn modern |applications into a 16-bit address space is very expensive. This drives |up software costs for users. Developers have to develop for two platforms. Who is forcing the developers to develop for the 286? The people who own them, obviously, with all the money they are waving at the developers. If the 286 owners think it's more economical to get software for their hardware instead of new software on new hardware, then that's their decision to make. |Sure, you can save a couple hundred on a 286. But what is that compared |to the cost of the ENTIRE system - hardware AND software. Peanuts. I think the question was, what does buying a 386 get you? I am pointing out that there is lots of life left in the 286. I don't like to see false claims for 386 systems (like WfW running better on a 386) go unchallenged. Sure there are people for whom the 386 is the right choice. I just want to expose all the pertinent facts. |Phil. Please throw that old junker on the trash heap. People like you If you are referring to what I personally use and own, you don't know. And it is irrelevant. Your choice and my choice is not the right choice for everyone. The proper thing to do is explain all the issues involved to the potential buyer. -- Phil Ngai, phil@amd.com {uunet,decwrl,ucbvax}!amdcad!phil The Microsoft Mouse is the only mouse worth having on a PC.
wcurtiss@x102c.harris-atd.com (Curtiss WC 67625) (03/10/90)
Ok, how about this: Buy a 286 if you want a fast XT. Everything made for the XT will run fine on it. Buy a 386(SX/DX) if you want multi-tasking/processing and run 386 software on it. When I can move from an XT to an AT for only a few hundred dollars and get a 10-times increase in speed, its worth it. Eventually the price differences should make this comparision more logical (i.e. the spread between a 8088/86 and 286 will drop and the gap between the 286 and 386SX/DX will widen). Not everyone needs a 386, afterall, 6502 based machines are still being sold. The moral of the story: everybody's needs are different, so know what you plan to do with your system and buy appropriately. You don't need 32 bits to balance your check book. ------------------------------------------------------------------------------- William Curtiss 407/984-6383 | - Standard disclamers apply - Harris GISD, Melbourne, FL 32902 | Internet: wcurtiss%x102c@ess.harris.com | <-- New address
cs4g6ag@maccs.dcss.mcmaster.ca (Stephen M. Dunn) (03/10/90)
In article <29405@amdcad.AMD.COM> phil@pepsi.AMD.COM (Phil Ngai) writes: $In article <8681@rosevax.Rosemount.COM> marknew@rosevax.UUCP (Mark Newman) writes: $|From Raymond Chen (raymond@math.berkeley.edu): $| There is no way to switch from protceted mode to real mode on a 286. $Of course not. What good is a protected mode if a random program $can violate it? Nobody ever said anything about allowing random chips to do it ... one of the big parts of having multitasking support is the provision of different privilege levels to different processes. The 80286, I believe, has four such levels. The privilege level of a process determines such things as what opcodes it can execute (for example, I/O is usually a no-no for low-privilege processes). By having an instruction at the highest privilege level that allows a switch back to real mode, random programs couldn't execute it (attempts to execute privileged instructions by non-privileged processes generate an interrupt which the OS has to handle). $|From Dean William Wallwey (wallwey@snoopy.colorado.edu): $| The 386 and 386SX both have very powerful memory mapping powers which $| are used by the new operating systems. $What new operating systems are you thinking of? Unix, perhaps? Possibly ... perhaps also the forthcoming 386-only version of OS/2. And if you call an EMS driver part of the operating system (and there are some grounds for making such a statement), that's another example. $| calls Virtual 86 mode so you can multitask DOS programs using Windows $| or DesqView for the 386. The small amount of money you save by buying $I still don't believe the DV386 uses virtual 86 mode. Does anyone have $any definitive data? Maybe I'll ask QD. I don't use DV or Windows, but it would make no sense to have a 386- specific version of a multitasking OS that allowed you to run DOS programs but didn't use the virtual 86 mode. $| The 386,386SX and 486 will $| run OS/2, mulitasking DOS, and UNIX, MUCH MUCH better than the 286 $| ever will. Even the new Windowing word processors from MicroSoft $Unix does run better on 386 and 486. I don't know what you mean by $multi-tasking DOS. It is quite possible to write an operating system, running on the 386, that uses multiple virtual 8086s and identical function calls to DOS which allows DOS program to be multitasked. There are already programs out there which do this with varying amounts of success. $| Even if you don't think that you'll need the power of the 386 group, $| you will soon see even the simple "home" programs like word pro- $| cessing and spreadsheets longing for the 386/386SX. (I guess they $| already do with WORD for WINDOWS, and Lotus 1-2-3 ver 3.0 even though $| they will "run" on a 286). $I don't think you know what you are talking about. Maybe you don't ... have you ever tried using a program like 1-2-3 R3 on a 286? Even on a 12 MHz 286, the program is a pig. From what I've heard, Word for Windows is also a pig. There are two distinct machine configurations for software: the absolute minimum system below which the program will not run at all, and the minimum practical system below which the program will run painfully slowly. For 1-2-3 R3, for example, the absolute minimum is a 286 with 1M of conventional + extended memory and a hard disk. The minimum practical system would have at least 2M of RAM and a processor running at 20 MHz or more. (Note that I'm talking DOS here ... 1-2-3 R3 for OS/2 absolutely requires something like 3 or 4M of memory). $With Windows 3.0, which does run in protected mode on a 286, due to $come out early this summer, you should consider exactly what the $extra cost of a 386 will buy you. Hold on a minute here ... what's the extra cost of a 386? For a 386SX system, we're talking a couple of hundred dollars difference between a 12 MHz 286 and a 16 MHz 386SX (more if you don't have a clue where to buy your system). For this you get the ability to run 386 software (which will become more prevalent in the future) and about 20-25% more speed. Not bad for the price. $|From John Limpert (johnl@n3dmc.UU.NET): $| In the future $| there will be a lot of software that will only be available for 80386 $| class machines. $Again, there is a very large market for 286 applications and they $simply will not be left unsupported. Nobody's saying that ... what people _are_ saying here, if you'd care to listen, is that software is being written that takes advantage of the 386, and more software will be written that way in the future. Nobody has said "Two years from now, all software will require a 386", because that's ludicrous. What will happen, though, is that more and more software will either take advantage of a 386 if you have one (like PKZIP), or will be written specifically to run on a 386 (you want examples? Look through a copy of BYTE. You'll see operating systems, compilers, database products - all applications where high performance is crucial, and all available in versions that require a 386). $Certainly Unix is one program that benefits from a 386. But users $don't buy PCs to run Unix, they buy PCs to run applications, and $Unix is one of the worst systems to try to buy applications for. $What kind of spreadsheets can you get? Anything that comes even $close to Excel? I hear you can finally get MS Word 5.0 on Unix. $A character based application, when DOS has Word for Windows. If $you've got the Unix religion, I guess you'll get a 386. But if $you want to get work done, you'll stay away from Unix. Well, you're perfectly entitled to voice your opinions on what operating system you want to use, what applications you want to use, and whether you want character-based or graphical interfaces. Just don't delude yourself into believing that everyone else has the same needs or desires that you do. -- Stephen M. Dunn cs4g6ag@maccs.dcss.mcmaster.ca <std_disclaimer.h> = "\nI'm only an undergraduate!!!\n"; **************************************************************************** "Oooh yes I need some love" - Rush
rob@mirror.tmc.com (Rob Limbert) (03/10/90)
In article <38299@iuvax.cs.indiana.edu> sl197009@silver.ucs.indiana.edu (Chima Echeruo) writes: >I have read many "experts" saying that the 286 is far slower than a 386. >I have a 20Mhz 286 and I am yet to find a DOS computer in my college that >performs faster than the 286 running DOS. >I have read in Byte that there are going to be versions of the 286 at 25 Mhz. >Is this true? >If it is, won't a 25Mhz 286 outperform an equivalent 386 while in real mode? Running DOS code, the 286 and 386 are about the same speed, clock-per-clock. 386 machines tend to have better memory architectures, and therefore have an edge. There will be 25mhz (and maybe faster) 286's eventually, and they'll run DOS code _very_ quickly if that's all you're looking for. Only a few 386's and 486's will be able to beat them at that task. >The only feature that bother me in the 286 is it's lack of "flat memory" and >also a lack of 286 specific software. The 'lack of flat memory', i.e. the 64K segment limit, always a headache, is an increasingly severe problem as applications and data grow in size and complexity. The 286's other major flaw is its inability to safely multitask real mode applications, or to easily mix real and protected mode code. When the 286 was designed, no one realized this would be a serious problem, but it's turned out to be a critical weakness. >When the 586 is ready for shipping by Intel, we will hear again how the 386 >is brain-damaged and not able to run the lastest sofware that is emerging. Not necessarily. Recall that the 386 has been around for about 4 years now, and no one's calling it 'brain-dead' yet. By the time the 286 had been out even 2 years, people were tossing that and various other epithets at it. You're assuming that the 286's only trouble is that it's not the latest, greatest chip on the block. That's partly true, but the 286 has numerous flaws that age alone can't explain. If age were the only problem, every chip would get hit with the 'brain dead' (or something similar) label eventually. Among the 8086 line (and even throwing in the 68000 line), the 286 has that title to itself.
phil@pepsi.amd.com (Phil Ngai) (03/10/90)
In article <38299@iuvax.cs.indiana.edu> sl197009@silver.ucs.indiana.edu (Chima Echeruo) writes: | |I have read many "experts" saying that the 286 is far slower than a 386. |I have a 20Mhz 286 and I am yet to find a DOS computer in my college that |performs faster than the 286 running DOS. | |I have read in Byte that there are going to be versions of the 286 at 25 Mhz. |Is this true? Yes. -- Phil Ngai, phil@amd.com {uunet,decwrl,ucbvax}!amdcad!phil The Microsoft Mouse is the only mouse worth having on a PC.
phil@pepsi.amd.com (Phil Ngai) (03/10/90)
In article <17995@boulder.Colorado.EDU> wallwey@boulder.Colorado.EDU (WALLWEY DEAN WILLIAM) writes: |Sometimes a 286 can beat a 386SX at the same speed, or obviously at |a faster rate, but is the extra approx 5% speed difference worth wasting |your future computer investment over? In case you haven't noticed, there's an entire industry dedicated to making an investment in microprocessors one of the worst ones around. As long as a 286 will run the applications you want for the next two or three years, it will probably be the best use of your money. Especially now that software like Windows 3.0 is coming out and you can finally use the 286's real power. |Also it's lack of Virtual '86 mode allowing true multitasking of real |mode DOS applications, and lack of special memory mapping |features. DESQview on a 286 with EEMS does multitasking in a very useful way. -- Phil Ngai, phil@amd.com {uunet,decwrl,ucbvax}!amdcad!phil The Microsoft Mouse is the only mouse worth having on a PC.
phil@pepsi.amd.com (Phil Ngai) (03/10/90)
In article <9982@portia.Stanford.EDU> dhinds@portia.Stanford.EDU (David Hinds) writes: |The 286 |also does not support memory interleaving systems, so for >16MHz, you will |be forced to have wait states on main memory. A 20MHz 386 can still work at You can, and people have, built 286 systems with cache. There is nothing special about a 286 that prevents you from using cache with it. You can, and people have, also built 286 systems with memory interleaving. -- Phil Ngai, phil@amd.com {uunet,decwrl,ucbvax}!amdcad!phil The Microsoft Mouse is the only mouse worth having on a PC.
phil@pepsi.amd.com (Phil Ngai) (03/10/90)
In article <17995@boulder.Colorado.EDU> wallwey@boulder.Colorado.EDU (WALLWEY DEAN WILLIAM) writes: |As for as this thread is concerned, the whole point of this disccussion |is whether you should buy a 286 or a 386 if you are buying a new system. |Again, the extra $100-$200 dollars spent on a 386SX clone (or better) |running at 16MHz around $900-$1100 compared with a 12Mhz 286 clone, |around $700-$900 is VERY WELL SPENT, and peanuts compared to costs you |will incure while owning a computer system!!!!!!!!!!! Maybe you're not aware of it, but the supply of 386 devices is getting very tight and since the producer has a monopoly, prices are sky-rocketing and there is no relief in sight. You better check on the CURRENT price differential before deciding that a 386 is a "good investment". -- Phil Ngai, phil@amd.com {uunet,decwrl,ucbvax}!amdcad!phil The Microsoft Mouse is the only mouse worth having on a PC.
wallwey@boulder.Colorado.EDU (WALLWEY DEAN WILLIAM) (03/10/90)
In article <3298@trantor.harris-atd.com> wcurtiss@x102c.ess.harris.com writes: >... the spread between a 8088/86 and 286 will >drop and the gap between the 286 and 386SX/DX will widen). I completely disagree... The price gap between the 286 and the 386SX/DX is going to get closer and closer--Just like it has been for 2 years!!!! There are also some new reasons why the price of the 386SX/386DX is even going to get closer. The new presure is the 486. You can already get complete 25Mhz 486 systems with 40MB drives and mono VGA for under $5000. This is and is going to further drive down the cost of the 386DX systems. This in turn presses down the cost of 386SX systems. This would press down the cost 286 systems, but they are already almost sold at COST prices. Also 8086 motherboards haven't gone down in 2 or more years!!! It's only the drop in price of external extras (cases, adapters, power supplies, monitors, disk drives) that have actually been driving prices of 8088 and 286 systems down in price in the last six months, and these apply just as well to 386SX/DX systems!!! Keep looking for 386 systems to become better deals, and 286 systems to look like more a waste of money. Dean
phil@pepsi.amd.com (Phil Ngai) (03/10/90)
In article <206900171@prism> rob@prism.TMC.COM writes: | Also the various DOS multitaskers like Windows 386, VM/386, and so on. DESQview 286 on a machine with EEMS seems to work about as well as DV 386. (I've tried both) The virtual 8086 mode may in principle make multitasking real mode processes easy but in practice it doesn't seem to make a whole lot of difference in products that have been delivered at this time. Neither Windows 386 or VM/386 are without their own problems, I hear. |>Windows, on the other hand, has a lot of software. And when 3.0 |>comes out, it will run in protected mode, allowing 16 megabytes of |>memory, ON A 286. | | It will still run better (offering multitasking of DOS applications |and better virtual memory support) on a 386. Have you seen this or are you assuming it? None of the "previews" of Windows 3.0 that I've seen have mentioned this. |The problem is that 286 protected mode is far |less versatile and powerful than 386 protected mode. That's why there's If you mean it doesn't have a large flat address space, that's true. But for multi-tasking, the 286 has plenty of potential for running protected mode programs. |I agree it's unfortunate |that the 286 hasn't been used more fully, but at this point, that's not |very relevant. It's quite relevant if you are interested in what software could be written for the 286. The fact that it may be painful is made up for by the incredible market that exists. |And the extra cost of a 386, especially |if you're willing to buy a non-name brand, is pretty marginal. That's becoming less and less true as the shortage of 386s gets worse. And the 386 market is a true monopoly. |Given the narrowing cost difference between 286 and 386 machines, the group |of tasks for which the 286 is the CPU of choice is shrinking, and will |continue to do so. On the contrary, I think that with software like Windows 3.0 coming out, the type of tasks which the 286 does well is going to increase dramatically. -- Phil Ngai, phil@amd.com {uunet,decwrl,ucbvax}!amdcad!phil The Microsoft Mouse is the only mouse worth having on a PC.
jca@pnet01.cts.com (John C. Archambeau) (03/11/90)
sl197009@silver.ucs.indiana.edu (Chima Echeruo) writes: >I have read many "experts" saying that the 286 is far slower than a 386. >I have a 20Mhz 286 and I am yet to find a DOS computer in my college that >performs faster than the 286 running DOS. > >I have read in Byte that there are going to be versions of the 286 at 25 Mhz. >Is this true? I've heard rumors that either AMD or Harris is going to produce a 25 MHz 286. The 286 does have a function in the world and will work with 286 based *nix operating systems. The reason that the 286 runs faster than the 386 on a MHz to MHz comparison is simple. The 286 has a smaller instruction set, but conversely, the 286 can't do the things the 386SX, 386, and 486 can. If you want raw speed and know 125% that you will never need the capabilities of a 386 then a 286 will work just fine. But the speed difference between a 286 and a 386 on a MHz to MHz comparison is very small. If you're unsure if you will need a 386 or 486 later then get a 286 that's upgradeable to a 386SX or 486 such as an ALR PowerFlex. To my knowledge those are the only machines that are upgredeable from a 286 to its younger and more powerful successors. That is the route I'm taking, I'm dumping my 16 MHz 286 for an ALR PowerFlex and plopping a 386SX board in it. I don't need the full 32-bit capabilities of the 386 (yet), just something to run 386 software. >If it is, won't a 25Mhz 286 outperform an equivalent 386 while in real mode? >The only feature that bother me in the 286 is it's lack of "flat memory" and >also a lack of 286 specific software. Yes, a 25 MHz 286 will out gun a 386 in raw computing speed. But remember, the 386 can do things the 286 can never do such as run multiple DOS sessions. Try running DOSMerge with a 286 based *nix system sometime, it crashes everytime. >I do not think that the 286 is dead, maybe it is not the state-of-the-art in >computer technology but it provides a very *fast* system for the average >computer user. If the SX does not go past 16Mhz, I do not see the benefit for >upgrading to a 386sx. Simple, to run 386 based software. With a 386SX you can run 386 operating systems and the whole works, of course you will be sacrificing some performance because of the bus being only 16-bits wide, but you are running things such as DesqView 386, SCO Xenix 386, Concurrent DOS 386, etc. That's the function of the 386SX, to provide 386 compatability at an almost 286 price. >When the 586 is ready for shipping by Intel, we will hear again how the 386 >is brain-damaged and not able to run the lastest sofware that is emerging. Not really, the 386 handles *nix based operating systems just fine. I see the 386 making *nix the new standard OS and if you want your DOS applications, you can still have them with VP/ix. The 386 is playing a significant role in the determination of what the next standard OS will be. DOS is about to head out the door, or maybe it will still be used, but in a manner such as VP/ix. Your OS will be able to run multiple DOS sessions. Everything else after the 386 is just a better mouse trap. >When MicroSoft came out with OS/2, PC/XT owners were advised to can their >machines and lay down the cash for a 286. After three years not much has been I wouldn't run OS/2 on my 286 if you paid me. Xenix takes up less memory than OS/2 which is a memory hog. Geez, just the word OS/2 makes me nauseous. // JCA /* **--------------------------------------------------------------------------* ** Flames : /dev/null | My opinions are exactly that, ** ARPANET : crash!pnet01!jca@nosc.mil | mine. Bill Gates couldn't buy ** INTERNET: jca@pnet01.cts.com | it, but he could rent it. :) ** UUCP : {nosc ucsd hplabs!hd-sdd}!crash!pnet01!jca **--------------------------------------------------------------------------* */
cs4g6ag@maccs.dcss.mcmaster.ca (Stephen M. Dunn) (03/11/90)
In article <38299@iuvax.cs.indiana.edu> sl197009@silver.ucs.indiana.edu (Chima Echeruo) writes:
$I have read many "experts" saying that the 286 is far slower than a 386.
$I have a 20Mhz 286 and I am yet to find a DOS computer in my college that
$performs faster than the 286 running DOS.
At the same clock speed, a 286 is slightly faster than a 386 (at least
when running DOS software); however, the fact that 386s are available that
run at significantly higher clock speeds than the current fastest 286 means
that you can buy a 386 system that will beat the performance of even the
best 286 machine.
$I have read in Byte that there are going to be versions of the 286 at 25 Mhz.
$Is this true?
Yes ... Harris, who has been producing 20 MHz 286s for a while now, has
just started making 25 MHz parts, and AMD plans to follow suit. Intel, who
is trying to kill the 286 since other companies are also making it, is not
likely to up the speed on their part. Instead, they are concentrating on
the 386SX to try to take over the current 286 market, and also the 386DX
and 486 for the high-performance market.
$If it is, won't a 25Mhz 286 outperform an equivalent 386 while in real mode?
A qualified yes. There will be exceptions, such as PKZIP, which sense
the presence of a 386 and utilize its instructions to achieve a significant
performance improvement. However, these exceptions are rare.
$I do not think that the 286 is dead, maybe it is not the state-of-the-art in
$computer technology but it provides a very *fast* system for the average
$computer user. If the SX does not go past 16Mhz, I do not see the benefit for
$upgrading to a 386sx.
It depends on what you want to do. There is some software that is written
specifically for the 386 family; if you want to run this software (such as
OS/2 version 2 when it arrives, and many products which have been on the
market for a while now), you have no choice. Also, don't forget that Intel
is expected to release a 20 MHz 386SX soon.
$When the 586 is ready for shipping by Intel, we will hear again how the 386
$is brain-damaged and not able to run the lastest sofware that is emerging.
It depends on what features they put into it. Two years ago, you might
have foretold the end of the 386 based on the fact that the 486 would be
coming out in a year or so ... but as it turned out, the 486 offers no
new operating modes. In fact, the 486 can be viewed as a fast 386+387
combination with a cache. It poses no threat to the 386 in terms of
operating modes, memory addressing capabilities, etc ... only in terms
of raw speed.
$When MicroSoft came out with OS/2, PC/XT owners were advised to can their
$machines and lay down the cash for a 286. After three years not much has been
In the same BYTE issue (March 1990) you mentioned, there is a short
article on DOS, Unix and OS/2 discussing such topics as why OS/2 hasn't
had the impact that it seemed it might have two years ago. They mention
such factors as increasing RAM prices (and we all know one feature of OS/2
is the ability to utilize, and the requirement for, more memory), a
somewhat incompatible DOS Compatibility Box, and the delay in the release
of the Presentation Manager.
--
Stephen M. Dunn cs4g6ag@maccs.dcss.mcmaster.ca
<std_disclaimer.h> = "\nI'm only an undergraduate!!!\n";
****************************************************************************
"So sorry, I never meant to break your heart ... but you broke mine."cs4g6ag@maccs.dcss.mcmaster.ca (Stephen M. Dunn) (03/11/90)
In article <9982@portia.Stanford.EDU> dhinds@portia.Stanford.EDU (David Hinds) writes:
$ [...] A 20MHz 386 can still work at
$"near 0 wait states" with interleaved memory. At higher speeds, a static
$RAM cache is necessary for avoiding wait states. I don't know if there are
$cache controllers for the 286, but I'm pretty sure the Intel cache controller
$only works with 386's.
I can't comment on what systems Intel's cache controller works with,
but don't forget that there were 386s with caches before Intel came up with
the cache controller. These were designed and implemented as custom
silicon, and the same can be done for the 286. Of course, it's so much
easier to use a ready-made cache controller ...
--
Stephen M. Dunn cs4g6ag@maccs.dcss.mcmaster.ca
<std_disclaimer.h> = "\nI'm only an undergraduate!!!\n";
****************************************************************************
"So sorry, I never meant to break your heart ... but you broke mine."marknew@rosevax.Rosemount.COM (Mark Newman) (03/12/90)
Greetings,
As the original poster of this thread I offer my apologies for starting
a processor war. I really didn't mean to. Honest! My aim was to tap the
collective intelligence of the net (no cheap jokes please) to help me
decide what processor my home PC should be based on.
But as long as the war is on, I'll offer my opinion: I've decided to
buy a 12 Mhz 286, for the following reasons:
1) This is a PC for the home.
2) It will not be running Unix or OS/2, or any other multitasking
environment (unless you count Windows/286 as multitasking).
3) The 286 has plenty of power for the applications I'll be running
(word proc., basic spreadsheet, communications, games, etc.)
4) The 286 still has quite a bit of life left and quite a lot of
software available.
In summary, the 286 fits my needs with plenty of power and capacity
available. Yes, it sure would be nice to have a SX. It would be even
nicer to have a 33 MHz 386. But I keep coming back to the question,
what for? Speed? 386 software? I won't be needing either *at home*.
One last thing. I went to the local Northgate factory outlet and asked
why they don't sell a 386SX machine. The answer was that they see the SX
as a stopgap chip, something that will not be long on the market because
it doesn't offer any useful performance over a 286 and yet isn't a full-
power 386 because of the 16-bit bus. Hmmm, this sounds familiar. I hear
the words "brain-dead" coming sometime for the SX.
Regards,
Mark Newman
marknew@rosevax.rosemount.comrob@mirror.tmc.com (Rob Limbert) (03/13/90)
In article <29449@amdcad.AMD.COM> phil@pepsi.AMD.COM (Phil Ngai) writes: >In article <206900171@prism> rob@prism.TMC.COM writes: >| Also the various DOS multitaskers like Windows 386, VM/386, and so on. >DESQview 286 on a machine with EEMS seems to work about as well as >DV 386. (I've tried both) The virtual 8086 mode may in principle make >multitasking real mode processes easy but in practice it doesn't seem >to make a whole lot of difference in products that have been delivered >at this time. Neither Windows 386 or VM/386 are without their >own problems, I hear. I've used both versions of DESQview, and DV 386, while not perfect, is distinctly better than DV 286 (at least for my purposes). I've used Windows 386 and VM/386 with good results. Most real-mode multitaskers are, in my opinion, so unstable as to be of little use. V86 multitaskers offer enough extra robustness and capability to make them worthwhile. V86 is no more a guarentee of stable multitasking than regular protected mode. Each theoretically permits completely safe multitasking, but it's possible, usually due to an OS bug, to crash a machine under protected or V86 mode. It's just much more difficult for an application level program to do it, by accident or on purpose. >|>Windows, on the other hand, has a lot of software. And when 3.0 >|>comes out, it will run in protected mode, allowing 16 megabytes of >|>memory, ON A 286. >| >| It will still run better (offering multitasking of DOS applications >|and better virtual memory support) on a 386. >Have you seen this or are you assuming it? None of the "previews" of >Windows 3.0 that I've seen have mentioned this. Windows 3.0 isn't an officially announced product yet, so discussing it isn't entirely appropriate. But I stand by the statement, which I wouldn't if it were just speculation. >|The problem is that 286 protected mode is far >|less versatile and powerful than 386 protected mode. That's why there's >If you mean it doesn't have a large flat address space, that's true. >But for multi-tasking, the 286 has plenty of potential for running >protected mode programs. There are numerous other advantages to 386 protected mode (paging, I/O permission mapping, the ability to mix real and protected mode tasks, 32-bit operations, etc). I'll grant, though, that the 286 has more multitasking potential than many people realize. This is a key point with regard to the original 'brain-dead 286' question. By many measures (and considering its age), it's a capable chip, with decent memory management and good performance. But in a world where DOS compatibility is everything, a world which the 286 designers understandably didn't foresee, it has real problems unless you're content to use it in real mode. And the 64K segment limit has gone from being a nuisance to a severe problem, as memory prices drop and programs grow in size and complexity. >|I agree it's unfortunate >|that the 286 hasn't been used more fully, but at this point, that's not >|very relevant. >It's quite relevant if you are interested in what software could >be written for the 286. The fact that it may be painful is made >up for by the incredible market that exists. I'm not saying that software developers should ignore the 286; given the size of the 286 market, that'd be unwise. I do maintain that, at this point in time, buying a 286 machine is an increasingly questionable decision unless you're not planning on keeping it for long, or unless you're certain that you won't want to run 386 software - which there's a growing volume of - in the foreseeable future. >|And the extra cost of a 386, especially >|if you're willing to buy a non-name brand, is pretty marginal. >That's becoming less and less true as the shortage of 386s gets >worse. And the 386 market is a true monopoly. You may be right; only time will tell. This could lead down a completely different rathole about second-sourcing. But, as machines are shipped with more and more memory, costlier video adapters, bigger hard disks, etc., the CPU itself becomes a progressively smaller factor in the overall system cost. Based on a recent look through PC Magazine, the cost difference between comparable 286 and 386 systems is as little as $150. >|Given the narrowing cost difference between 286 and 386 machines, the group >|of tasks for which the 286 is the CPU of choice is shrinking, and will >|continue to do so. >On the contrary, I think that with software like Windows 3.0 coming >out, the type of tasks which the 286 does well is going to increase >dramatically. True, but it will also dramatically increase the type of tasks which the 386 does better. OS/2 version 2.0 will do the same thing. Maybe underneath it all we're both bemoaning the same problem - the large time lag between hardware and software. In the Intel world, that lag is several years, and it's condemned the 286, 386, and 486 to be run mostly as fast 8088s. All these chips, including the 286, have lots of untapped potential. My argument is that the 386 has much more than the 286 (and that potential has already been more fully realized), and, given the relatively small price difference, is a wiser purchase for most purposes.
sl197009@silver.ucs.indiana.edu (Chima Echeruo) (03/13/90)
>In article <9982@portia.Stanford.EDU> dhinds@portia.Stanford.EDU (David Hinds) writes: >$ [...] A 20MHz 386 can still work at >$"near 0 wait states" with interleaved memory. At higher speeds, a static >$RAM cache is necessary for avoiding wait states. I don't know if there are >$cache controllers for the 286, but I'm pretty sure the Intel cache controller >$only works with 386's. My 20Mhz 286 which is based on the Chips & Tech NEAT chipset claims ZERO wait state performance using interleaved memory. Norton SI = 22.5 ------------- Chima Echeruo sl197009@silver.ucs.indiana.edu -------------------------------
phil@pepsi.amd.com (Phil Ngai) (03/13/90)
In article <25F7F56B.11734@maccs.dcss.mcmaster.ca> cs4g6ag@maccs.dcss.mcmaster.ca (Stephen M. Dunn) writes: | Maybe you don't ... have you ever tried using a program like 1-2-3 R3 |on a 286? Even on a 12 MHz 286, the program is a pig. From what I've |heard, Word for Windows is also a pig. There are two distinct machine But that doesn't have anything to do with 286 or 386. WfW doesn't (I believe) do anything special on a 386. I don't think 123 R3 does either. So getting a 386 for these programs is pointless. |$With Windows 3.0, which does run in protected mode on a 286, due to |$come out early this summer, you should consider exactly what the |$extra cost of a 386 will buy you. | | Hold on a minute here ... what's the extra cost of a 386? For a |386SX system, we're talking a couple of hundred dollars difference between |a 12 MHz 286 and a 16 MHz 386SX (more if you don't have a clue where to |buy your system). For this you get the ability to run 386 software |(which will become more prevalent in the future) and about 20-25% more |speed. Not bad for the price. A SX tends to be a little slower than a 286 running at the same speed. You can buy 286s running at 20 MHz now and they will be running at 25 MHz soon. A 386DX does run faster but then you have to pay a rather hefty premium for the speed. 386 software will be more prevalent in the future but so will 286 software in the form of Windows 3.0 and all the associated applications. | Well, you're perfectly entitled to voice your opinions on what |operating system you want to use, what applications you want to use, |and whether you want character-based or graphical interfaces. Just |don't delude yourself into believing that everyone else has the same |needs or desires that you do. I do think there is a place for 386s, I just think a lot of people are selling the 286 short for some vague benefit in the indefinite future and are forgetting that an investment in microprocessors is a lousy investment. -- Phil Ngai, phil@amd.com {uunet,decwrl,ucbvax}!amdcad!phil The Microsoft Mouse is the only mouse worth having on a PC.
wallwey@boulder.Colorado.EDU (WALLWEY DEAN WILLIAM) (03/13/90)
In article <25F9E2B6.20389@maccs.dcss.mcmaster.ca> cs4g6ag@maccs.dcss.mcmaster.ca (Stephen M. Dunn) writes: > In the same BYTE issue (March 1990) you mentioned, there is a short >article on DOS, Unix and OS/2 discussing such topics as why OS/2 hasn't >had the impact that it seemed it might have two years ago. They mention >such factors as increasing RAM prices (and we all know one feature of OS/2 >is the ability to utilize, and the requirement for, more memory), a >somewhat incompatible DOS Compatibility Box, and the delay in the release >of the Presentation Manager. >-- >Stephen M. Dunn cs4g6ag@maccs.dcss.mcmaster.ca > <std_disclaimer.h> = "\nI'm only an undergraduate!!!\n"; >**************************************************************************** > "So sorry, I never meant to break your heart ... but you broke mine." Actually most of those reasons, have changed, or will very soon! Memory prices have droped to "low levels" again. Vers 2.0 will have a much more compatable "DOS Compatibility Box" that is able to actually multitask DOS applications as well. (This is because Version 2.0 will use the Virtual Machine mode of the 386 and hence will require a 386 for this and other reasons!!!) And Finally the new ver 1.2 Presentation Manager is even better looking than UNIX GUIs (like Motif!) not to mention much faster on an equivalent machine!) Another thing to keep in mind, most of the big players are spending considerable time, money, effort, and much of their R&D on OS/2, like MicroSoft, Lotus, AutoCad, WordPerfect corp. and others. Most of them have announced products that make their DOS counterparts sound like they are from the Dark Ages! Dean Wallwey
Duel@cup.portal.com (Omid M Farr) (03/13/90)
After reading some debate on this topic of the 286 vs. 386, I've come to wonder what is the advatange of the 486 today? I don't think we can expect much software to come out for the 486 specifically this year, so is there really an advantage to owning and using a 486? If a 386 runs software made for the 286 (assuming they are at the same clock rate) at similar speeds, then I take it the 386/486 would follow the same pattern. So this is what I think is going on, Please correct me if I am wrong... A 25 MHz 486 can run software written for the 386 specifically, but a 25 MHz 386 would run the same program at about the same speed, meaning a 486 really doesn't (today) have any advantages unless you intend on writing software for it. Is this true? Or does the rule apply only to 286/386? Thanks.
marshall@wind55.seri.gov (Marshall L. Buhl) (03/14/90)
phil@pepsi.amd.com (Phil Ngai) writes: >In article <25F7F56B.11734@maccs.dcss.mcmaster.ca> cs4g6ag@maccs.dcss.mcmaster.ca (Stephen M. Dunn) writes: >| Maybe you don't ... have you ever tried using a program like 1-2-3 R3 >|on a 286? Even on a 12 MHz 286, the program is a pig. From what I've >|heard, Word for Windows is also a pig. There are two distinct machine >But that doesn't have anything to do with 286 or 386. WfW doesn't (I >believe) do anything special on a 386. I don't think 123 R3 does either. >So getting a 386 for these programs is pointless. Phil, You're going to have to live with that ancient relic for at least 4 years. Just because there isn't that much 32-bit software today, doesn't mean that will continue to be true. My FORTRAN programs run nearly twice as fast in 32-bit mode compared to 16-bit mode. The 286 is a very OLD architecture. Look to the future. Quit recommending that people invest in this old stuff. I'm not saying to junk all those old systems, just don't buy any NEW 286s. Use the old ones in non-critical situations. People who insist in investing in ancient technology are holding the rest of us back. Remember folks. The CPU is a very small part of the cost of a COMPLETE system. When you pay more for a 386 system, you are getting faster memory, hard disks, etc. too - not just a faster processor. Also remember that 1-2-3 costs the same whether you will run it on a 8088 or a 25 MHz 80486. Your $2500 LaserJet costs the same too. Don't slow down all that nice stuff with a wimpy processor. If your time is valuable to you, then you can't AFFORD a 286 - it's just too expensive. 16-bit computers are living on borrowed time. 32-bit is the way to go for the near term. Of course, the 64-bit CPU is not far off. Maybe then, we will finally get the performance we need to run a GUI. -- Marshall L. Buhl, Jr. EMAIL: marshall@wind55.seri.gov Senior Computer Engineer VOICE: (303)231-1014 Wind Research Branch 1617 Cole Blvd., Golden, CO 80401-3393 Solar Energy Research Institute Solar - safe energy for a healthy future
phil@pepsi.amd.com (Phil Ngai) (03/14/90)
In article <1990Mar13.201259.11097@seri.gov> marshall@wind55.seri.gov (Marshall L. Buhl) writes: |You're going to have to live with that ancient relic for at least 4 |years. Just because there isn't that much 32-bit software today, doesn't First of all, a 286 is not an ancient relic. Microsoft is only now getting around to releasing a version of Windows which can take full advantage of the 286's power. I'm glad you admit there isn't much 32-bit software. Second, I think your 4 year figure is off. If 386 prices went way down next year, buying a 286 now and a 386 in 1991 could be cheaper than buying a 386 now. Just imagine the price war that will break out when a viable 386 second source comes along. |Quit recommending that people invest |in this old stuff. I think this is the root of your problem. You see a computer as an investment. I see them as a purchase to get a job done. An asset that depreciates incredibly fast. Microprocessors make a lousy investment, especially with the price war that will erupt when the 386 has a second source. |People who insist in investing in ancient technology are holding the |rest of us back. Now you're getting silly. You want the people who can least afford a high end system, the people who are looking at 286s, to spend more money for something that probably won't make much difference to them, because you think it will somehow help you? Even if that were true, which I doubt, it's a very selfish attitude. |Remember folks. The CPU is a very small part of the cost of a COMPLETE |system. When you pay more for a 386 system, you are getting faster |memory, hard disks, etc. too - not just a faster processor. And what if the customer doesn't need all this horsepower? The market is made up of very different people. Not everyone is like you. It's not much different than you telling everyone to get a 600 megabyte hard disk because they might need it someday. -- Phil Ngai, phil@amd.com {uunet,decwrl,ucbvax}!amdcad!phil The Microsoft Mouse is the only mouse worth having on a PC.
jwi@cbnewsj.ATT.COM (Jim Winer @ AT&T, Middletown, NJ) (03/14/90)
> Marshall L. Buhl writes: > > ... The 286 is a very OLD > architecture. Look to the future. Quit recommending that people invest > in this old stuff. I'm not saying to junk all those old systems, just > don't buy any NEW 286s. Use the old ones in non-critical situations. and always pay more for bells and whistles that you don't need, whether it's a car, a TV, a computer, or solar energy. > People who insist in investing in ancient technology are holding the > rest of us back. Then why don't we all have nuclear power plants -- seems to me that people said this 20 years ago too. > Remember folks. The CPU is a very small part of the cost of a COMPLETE > system. When you pay more for a 386 system, you are getting faster > memory, hard disks, etc. too - not just a faster processor. My 286/20 is pretty fast. It has 80 ns interleaved memory and 0 wait states. It is comparable in speed to a 386/20 if not faster. It will take the fastest disk I want to pay for. I don't need a 386 to get either fast memory, fast disk or a faster processor. The 286/25 machines are available now. > ... > 16-bit computers are living on borrowed time. 32-bit is the way to go for > the near term. Of course, the 64-bit CPU is not far off. Maybe then, we > will finally get the performance we need to run a GUI. Smarter computers can be run more easily by dumber people -- agreed. Why don't you learn to use what you have instead of complaining that others are holding you back? Frankly, I wouldn't want to hold you back -- the faster you fall off the edge, the less I have to hear you whining. Jim Winer -- jwi@mtfme.att.com -- Opinions not represent employer. ------------------------------------------------------------------
cs4g6ag@maccs.dcss.mcmaster.ca (Stephen M. Dunn) (03/15/90)
In article <29474@amdcad.AMD.COM> phil@pepsi.AMD.COM (Phil Ngai) writes: $In article <25F7F56B.11734@maccs.dcss.mcmaster.ca> cs4g6ag@maccs.dcss.mcmaster.ca (Stephen M. Dunn) writes: $| Maybe you don't ... have you ever tried using a program like 1-2-3 R3 $|on a 286? Even on a 12 MHz 286, the program is a pig. From what I've $|heard, Word for Windows is also a pig. There are two distinct machine $But that doesn't have anything to do with 286 or 386. WfW doesn't (I $believe) do anything special on a 386. I don't think 123 R3 does either. $So getting a 386 for these programs is pointless. I can't agree with that, Phil. I don't consider upgrading a CPU to be useful only if it lets you do things you could never do before. I also consider it useful if it speeds up the slower things you used to do. Perhaps your concept of pointlessness is different from mine, and you're quite happy to waste time with a slow program rather than get a computer with a faster CPU. Personally, I can't wait to get my hands on my new 386SX system (I'm currently using a 12 MHz 286), and it isn't because I want to run OS/2 V2.0 or anything like that. -- Stephen M. Dunn cs4g6ag@maccs.dcss.mcmaster.ca <std_disclaimer.h> = "\nI'm only an undergraduate!!!\n"; **************************************************************************** "So sorry, I never meant to break your heart ... but you broke mine."
rob@mirror.tmc.com (Rob Limbert) (03/15/90)
In article <38483@iuvax.cs.indiana.edu> sl197009@silver.ucs.indiana.edu (Chima Echeruo) writes: >>In article <9982@portia.Stanford.EDU> dhinds@portia.Stanford.EDU (David Hinds) writes: >>$ [...] A 20MHz 386 can still work at >>$"near 0 wait states" with interleaved memory. At higher speeds, a static >>$RAM cache is necessary for avoiding wait states. I don't know if there are >>$cache controllers for the 286, but I'm pretty sure the Intel cache >>$controller only works with 386's. >My 20Mhz 286 which is based on the Chips & Tech NEAT chipset claims ZERO >wait state performance using interleaved memory. Norton SI = 22.5 The key word here is 'claims'. It's common practice to advertise fast 286, 386, and 486 machines as having 0 wait states when they actually don't. Of course, the phrase 'zero wait state performance', which shows up in ads a lot, is vague enough that it doesn't mean much. Most of the more complex memory architectures (caching, interleaving, etc.) work by allowing 0 wait states under certain conditions, and >0 wait states under others. Thus the only really meaningful figure is an average for a supposedly typical code sequence. A good cache controller will offer 0 wait states perhaps 95% of the time, and 4 or 5 wait states (for cache misses) the rest of the time, for an average of around 0.2 wait states. A typical interleaved system might work with 0 wait states 50% of the time, and 2 wait states the other 50%, for an average of about 1. Running a 20Mhz or faster machine completely without wait states would require using very fast DRAM or SRAM for main memory, and would be prohibitively expensive. By the way, though Intel doesn't make a cache controller for the 286, there's nothing that prevents a 286 from being cached. 286 based accelerator boards have been using caches for years, and some faster 286 machines use caches.
ilan343@violet.berkeley.edu (03/15/90)
In article <29486@amdcad.AMD.COM> phil@pepsi.AMD.COM (Phil Ngai) writes: >First of all, a 286 is not an ancient relic. Microsoft is only now >getting around to releasing a version of Windows which can take >full advantage of the 286's power. I used a IBM-PC AT for the first time some time in 1984. This is is ancient in micro processor years. Microsoft is the one thet is kind of slow on this one (and on everything else they do).
phil@pepsi.amd.com (Phil Ngai) (03/15/90)
In article <25FE7267.24003@maccs.dcss.mcmaster.ca> cs4g6ag@maccs.dcss.mcmaster.ca (Stephen M. Dunn) writes: |In article <29474@amdcad.AMD.COM> phil@pepsi.AMD.COM (Phil Ngai) writes: |$But that doesn't have anything to do with 286 or 386. WfW doesn't (I |$believe) do anything special on a 386. I don't think 123 R3 does either. |$So getting a 386 for these programs is pointless. | | I can't agree with that, Phil. I don't consider upgrading a CPU to |be useful only if it lets you do things you could never do before. I |also consider it useful if it speeds up the slower things you used to |do. Now you're changing the subject of the discussion. The original question was, is it worth the extra money to get a 386? Some claimed that the hope of 32-bit software was worth spending extra money now. There were allusions to the capabilities of the mentioned programs to use a 386, which I questioned. If you simply want a faster clock rate processor, then that's a different question. | Personally, I can't wait to get my hands on my new 386SX system |(I'm currently using a 12 MHz 286), and it isn't because I want to |run OS/2 V2.0 or anything like that. Then why aren't you getting a 20 MHz 286 which is cheaper than a 16 MHz 386sx? Or even a 25 MHz 286? -- Phil Ngai, phil@amd.com {uunet,decwrl,ucbvax}!amdcad!phil Boycott the census! With the history of abuse census data has, do you really trust the government?
marshall@wind55.seri.gov (Marshall L. Buhl) (03/16/90)
phil@pepsi.amd.com (Phil Ngai) writes: >In article <1990Mar13.201259.11097@seri.gov> marshall@wind55.seri.gov (Marshall L. Buhl) writes: >|You're going to have to live with that ancient relic for at least 4 >|years. Just because there isn't that much 32-bit software today, doesn't >First of all, a 286 is not an ancient relic. Microsoft is only now >getting around to releasing a version of Windows which can take >full advantage of the 286's power. When I say ancient relic, I mean it was developed something like 7 to 10 years ago. In the computer business, anything more than 5 years old is ancient history. >Second, I think your 4 year figure is off. If 386 prices went way down >next year, buying a 286 now and a 386 in 1991 could be cheaper than >buying a 386 now. Just imagine the price war that will break out when >a viable 386 second source comes along. You missed the point. The processor is a very minor part of the expense of a TOTAL computer system. What would a cheaper 386 save - maybe $200? That's nothing. Also, I'm not talking about home PCs. I'm talking about those being used in business. Also, you can wait forever for cheaper prices. >|Quit recommending that people invest >|in this old stuff. >I think this is the root of your problem. You see a computer as an >investment. I see them as a purchase to get a job done. An asset that >depreciates incredibly fast. Microprocessors make a lousy investment, >especially with the price war that will erupt when the 386 has a >second source. I see them as an investment in a tool that I need to get the job done in a reasonable amount of time. I'm NOT recommending people go out and buy the lastest and greatest - not unless they really need it. The Compaq 386/16 came out 3.5 years ago. It's old hat now. I don't have any 386/33s or 486/25s. Just get something that will still be useful four years from now. A 286 won't. It will be well over a decade old by then. That's about 200 in computer years. 8-) And once again. I can get a 386 for just a few hundred dollars - qty 1. The 16 MHz chip probably doesn't cost more than $200 in quantity. Having a price war won't make that much difference, unless someone pays you to use their chip. 8-) >|People who insist in investing in ancient technology are holding the >|rest of us back. >Now you're getting silly. You want the people who can least afford >a high end system, the people who are looking at 286s, to spend >more money for something that probably won't make much difference >to them, because you think it will somehow help you? Even if that >were true, which I doubt, it's a very selfish attitude. Phil - I'm not talking about poor folks. I'm talking about businesses. Poor folks don't spend thousands of dollars on software. Businesses do. If the developers have to develop software for 16-bit computers along with 32-bit, it will slow down development and add costs. I'm not talking about games. I'm talking about DTP, CAD, Advanced WP, etc. Look at all the problems Lotus is having because they had to make two versions of 1-2-3. Home PC users can't afford 1-2-3, they get a shareware clone. 1-2-3 is for businesses. As far as your "won't make any difference to them" remark, I've NEVER (not even once) met anyone who used a 386 for a while who didn't HATE going back to a 286. Low-end users think of this as being spoiled. Others think of it as being able to get more work done in less time. >|Remember folks. The CPU is a very small part of the cost of a COMPLETE >|system. When you pay more for a 386 system, you are getting faster >|memory, hard disks, etc. too - not just a faster processor. >And what if the customer doesn't need all this horsepower? The market is >made up of very different people. Not everyone is like you. It's not >much different than you telling everyone to get a 600 megabyte hard disk >because they might need it someday. It's a hell of a lot cheaper to upgrade peripherals that systems. It's not the same at all. Whether you like it or not, the world is going GUI. A 386/25 is not powerful enough to run Windows/386 even today. And please don't get me wrong. Until last year, I have used character based software exclusively. That's since '72. I don't use Macs either. As a matter of fact, I have quit using Windows because my 386/25 isn't fast enough. I can get more work done with PC Word than Word for Windows. I hope to buy a 486/33 around the end of the year. Maybe then, I'll be able to run Windows without it holding me back. As you can tell from my age, I'm an old mainframer. I used to complain to my boss that having to wait 5 seconds for response was killing me. We have a running gag about me never having enough power. I contended that if I had to wait more than a second, then I lose my rhythm. One day he came to me with an article about response time. It reported on a study by IBM (?). Their study found that if a user had to wait more than a HALF second, then their minds started thinking about other things and productivity went down. Your mind just can't keep idling for very long - it needs to keep busy. Fast response is critical in a business enviroment where time is money. Slow computers are MORE expensive the fast computers. -- Marshall L. Buhl, Jr. EMAIL: marshall@wind55.seri.gov Senior Computer Engineer VOICE: (303)231-1014 Wind Research Branch 1617 Cole Blvd., Golden, CO 80401-3393 Solar Energy Research Institute Solar - safe energy for a healthy future
phil@pepsi.amd.com (Phil Ngai) (03/17/90)
In article <1990Mar15.202659.4744@seri.gov> marshall@wind55.seri.gov (Marshall L. Buhl) writes: |When I say ancient relic, I mean it was developed something like 7 to 10 |years ago. In the computer business, anything more than 5 years old is |ancient history. I don't care how many years something has been around, if it is still being used and its full potential hasn't been tapped yet, then it isn't an ancient relic. |You missed the point. The processor is a very minor part of the expense |of a TOTAL computer system. What would a cheaper 386 save - maybe $200? |That's nothing. Also, I'm not talking about home PCs. I'm talking about |those being used in business. Maybe you came in in the middle but I'm pretty sure the original poster (who has been overwhelmed by what he started) was asking for home use. Furthermore, he did decide to get a 286. Also your view of business use is quite narrow minded. It extends only to the tube in front of your face. If you ever looked at data on the software market, you'd find by far the biggest segments are things like Lotus 123 and Wordperfect, where 286s do just fine. For the manager with a room full of word processors or clerks, a savings of several hundred dollars per seat can be very important. |Also, you can wait forever for cheaper prices. That's a simplistic cliche from someone obviously uninterested in the details of the market he claims to dispense expert information on. |I see them as an investment in a tool that I need to get the job done |in a reasonable amount of time. I'm NOT recommending people go out Sorry Marshall, but the only ones who care about what you need to get your job done are you and your boss. As I understand your job, it's very unlikely that your recommendations are relevant to the entire PC market. |If the developers have to develop software for 16-bit computers along |with 32-bit, it will slow down development and add costs. I'm not I think your attitude about 16-bit processors is just like your attitude about guns. Idealistic and unrealistic. You think that if guns were illegal, then violence would be ended. You ignore the fact that there are hundreds of millions of guns out there and criminals won't stop using guns just because it's against the law. Do you really think you can stop developers from supporting 16-bit processors by telling everyone to buy a 386? Sorry, there's too many 286s out there already. Maybe we should make everyone register their "dangerous" 286s. Then next year, we could make them illegal and confiscate them. Save the world for 32-bit software. |Look |at all the problems Lotus is having because they had to make two versions |of 1-2-3. So who are you to tell Lotus what to do? If Lotus thinks they can make money this way, it's their business. If their customers think this is the best path for them, that's *their* business. Are you such an expert that you know better than everyone else the best way to do their job? |It's a hell of a lot cheaper to upgrade peripherals that systems. All you have to do is swap a motherboard. Even if you can't figure it out, you can pay someone a small amount of money to do it for you. |Whether you like it or not, the world is going GUI. ... |I have quit using Windows because my 386/25 isn't fast enough. I can |get more work done with PC Word than Word for Windows. I hope to buy a |486/33 around the end of the year. Maybe then, I'll be able to run Windows |without it holding me back. What are you saying here? That GUIs are too slow even for a 386 and that you need a 486 to use them? If the whole world is going GUI, then the whole world is getting 486s? You've fallen off the deep end. |As you can tell from my age, I'm an old mainframer. I used to complain to |my boss that having to wait 5 seconds for response was killing me. Then you better not use GUIs... Once again, you don't know what you're talking about if you think that your needs are typical or have anything to do with what's best for everyone else. Fortunately, you're not in the computer business or you'd go hungry very fast. -- Phil Ngai, phil@amd.com {uunet,decwrl,ucbvax}!amdcad!phil Boycott the census! With the history of abuse census data has, can you afford to trust the government?
jca@pnet01.cts.com (John C. Archambeau) (03/18/90)
phil@pepsi.amd.com (Phil Ngai) writes: >Then why aren't you getting a 20 MHz 286 which is cheaper than >a 16 MHz 386sx? Or even a 25 MHz 286? Cheaper my foot. Apparently you haven't been keeping up in the Computer Shopper. The 16 MHz 386SX motherboards are about up to par in price with the 20 MHz 286 boards. I view it as paying for 386 compatability which is what you need for 386 based Unix. Well worth the money since I'm fed up with DOS and want to pack the Sun SPARCstation 1 on my desk at work in my car and take it home for an indefinate period of time. But I've found a more legal solution to my desire for Unix, ESIX and a 386SX system. Works for me. If I want DOS, I can run VP/ix which won't work in a million years on a 286 based system. Unix will work on a 286, but you can't have virtual DOS sessions. Try DOSMerge on MicroPort Unix 286, it crashes quite nicely. // JCA /* **--------------------------------------------------------------------------* ** Flames : /dev/null | My opinions are exactly that, ** ARPANET : crash!pnet01!jca@nosc.mil | mine. Bill Gates couldn't buy ** INTERNET: jca@pnet01.cts.com | it, but he could rent it. :) ** UUCP : {nosc ucsd hplabs!hd-sdd}!crash!pnet01!jca **--------------------------------------------------------------------------* */
kdq@demott.COM (Kevin D. Quitt) (03/19/90)
The 486 takes advantage of RISC-type hardware. At the same clock speeds, the 486 shold be 2-3 times faster than a 386. This make it slightly more powerful than the 68030. kdq -- Kevin D. Quitt Manager, Software Development DeMott Electronics Co. VOICE (818) 988-4975 14707 Keswick St. FAX (818) 997-1190 Van Nuys, CA 91405-1266 MODEM (818) 997-4496 Telebit PEP last 34 12 N 118 27 W srhqla!demott!kdq kdq@demott.com "Next time, Jack, write a God-damned memo!" - Jack Ryan - Hunt for Red Oct.
wallwey@boulder.Colorado.EDU (WALLWEY DEAN WILLIAM) (03/20/90)
In article <27842@cup.portal.com> Duel@cup.portal.com (Omid M Farr) writes: >After reading some debate on this topic of the 286 vs. 386, I've come to >wonder what is the advatange of the 486 today? I don't think we can expect >much software to come out for the 486 specifically this year, so is there >really an advantage to owning and using a 486? Yes, and no.....SPEED!!!!! When Intel made the 386, they did it right-- as opposed to 286---Even Intel wishes the 286 didn't exist.....There are no new modes in the 486 over the 386--You are right in saying that no software will take advantage of the 486 because from the programming point of view there is none. Anyways there are two reasons why you would want to buy a 486---At the same speed it is cosiderably faster---It's internal structure resembles many RISC designs and the built in FPU is much faster than its 387 counterpart at the same speed! The second reason is cost---486 systems, when they become a commodity (probably late this year) will actually be cheaper to build than the High performance 386 systems they replace. This is because, not only is the 486 faster than its 386 counter parts, but also it includes 2 major EXTRAS you have to add in a 386 system for less. Not only does the 486 include the CPU, but it includes the CACHE and FPU! The cost of the 486 chip, about $1000 in qty., is easily cheaper than the 386, 387, cache controller, and cache memory it replaces! Not only that, but the motherboard can be simplier in a 486, just because of the smaller chip count. >If a 386 runs software made for the 286 (assuming they are at the same clock >rate) at similar speeds, then I take it the 386/486 would follow the same >pattern. >So this is what I think is going on, Please correct me if I am wrong... >A 25 MHz 486 can run software written for the 386 specifically, but a >25 MHz 386 would run the same program at about the same speed, meaning a 486 >really doesn't (today) have any advantages unless you intend on writing >software for it. > >Is this true? No. Just the opposite, the 486 is much faster at the same speed, but has very few (nill to most porgramers), differences in capabilities. >Or does the rule apply only to 286/386? There is a VERY BIG difference between the 386 and 286---The 386 and 486 are 32 bit machines, the 286 and its older brother, the 8086, are 16 bit machines!!!!!! Of course the 386/486 can run exactly the same software, and are backwards compatible with the 286 and 8086--meaning they can run the old (figuratively) 286 and 86 software. Dean Wallwey
kdq@demott.COM (Kevin D. Quitt) (03/20/90)
In article <18603@boulder.Colorado.EDU> wallwey@boulder.Colorado.EDU (WALLWEY DEAN WILLIAM) writes: > >In article <27842@cup.portal.com> Duel@cup.portal.com (Omid M Farr) writes: >>After reading some debate on this topic of the 286 vs. 386, I've come to >>wonder what is the advatange of the 486 today? I don't think we can expect >>much software to come out for the 486 specifically this year, so is there >>really an advantage to owning and using a 486? > >Yes, and no.....SPEED!!!!! When Intel made the 386, they did it right-- Actually, they just did it a little less wrong. >No. Just the opposite, the 486 is much faster at the same speed, but has >very few (nill to most porgramers), differences in capabilities. Intel has finally followed in the footsteps of Motorola, Dec, National, CDC, etc. (Presented in no particular order). kdq -- Kevin D. Quitt Manager, Software Development DeMott Electronics Co. VOICE (818) 988-4975 14707 Keswick St. FAX (818) 997-1190 Van Nuys, CA 91405-1266 MODEM (818) 997-4496 Telebit PEP last 34 12 N 118 27 W srhqla!demott!kdq kdq@demott.com "Next time, Jack, write a God-damned memo!" - Jack Ryan - Hunt for Red October
cs4g6ag@maccs.dcss.mcmaster.ca (Stephen M. Dunn) (03/21/90)
In article <18603@boulder.Colorado.EDU> wallwey@boulder.Colorado.EDU (WALLWEY DEAN WILLIAM) writes:
$Yes, and no.....SPEED!!!!! When Intel made the 386, they did it right--
$as opposed to 286---Even Intel wishes the 286 didn't exist.....There are
$no new modes in the 486 over the 386--You are right in saying that no
$software will take advantage of the 486 because from the programming
$point of view there is none.
Well, code generated with the 486 in mind can be tweaked to run 10-15%
faster than code not tweaked with the 486 in mind ... this isn't a new
programming feature, but it is affected by your programming (or your
compiler's optimization).
$Not only does the 486 include the CPU, but it includes the CACHE and
$FPU! The cost of the 486 chip, about $1000 in qty., is easily cheaper
$than the 386, 387, cache controller, and cache memory it replaces! Not
Yes, having that on-chip cache is wonderful ... but don't forget, it's
only an 8K cache, whereas cached 386 systems typically have 32, 64, or 128K
caches. So if you want the same size cache as you'd have on your 386,
you _will_ have to add it yourself. Still, an 8K cache is a big step
forward.
--
Stephen M. Dunn cs4g6ag@maccs.dcss.mcmaster.ca
<std_disclaimer.h> = "\nI'm only an undergraduate!!!\n";
****************************************************************************
"So sorry, I never meant to break your heart ... but you broke mine."