dhinds@portia.Stanford.EDU (David Hinds) (02/27/90)
In article <25E98AD7.9457@maccs.dcss.mcmaster.ca>, cs4g6ag@maccs.dcss.mcmaster.ca (Stephen M. Dunn) writes: > $For example the numeric coprocessor and a memory cache are include in the > $chip. As for upgrading, that depends on whether your 386 was designed > $for upgrading. There is a major difference in the motherboard design > [...] > > Yup ... the whole motherboard has to be replaced unless it was > designed for upgrading, and very few machines are designed for upgrading. > -- It seems like it would not be all that hard to design a teeny daughter board that fits into the 80386 socket, and carries an 80486. Is there a real difference in design of the support circuitry, that an 80486 wouldn't work in place of an 80386? From what I understand about the '486, to the outside world (i.e., the motherboard), it shouldn't look much different from an 80386. I would guess that the 80486 might have a couple extra control lines that might need to be tied off somehow, but it should be able to start up off of a standard 80386 BIOS, even. When '486 prices fall, this seems like a potentially inexpensive upgrade - instead of getting a fast 80387, for example. - David Hinds dhinds@portia.stanford.edu
v126gfl3@ubvms.cc.buffalo.edu (03/01/90)
In article <9589@portia.Stanford.EDU>, dhinds@portia.Stanford.EDU (David Hinds) writes: > In article <25E98AD7.9457@maccs.dcss.mcmaster.ca>, cs4g6ag@maccs.dcss.mcmaster.ca (Stephen M. Dunn) writes: >> $For example the numeric coprocessor and a memory cache are include in the >> $chip. As for upgrading, that depends on whether your 386 was designed >> $for upgrading. There is a major difference in the motherboard design >> [...] >> >> Yup ... the whole motherboard has to be replaced unless it was >> designed for upgrading, and very few machines are designed for upgrading. >> -- > It seems like it would not be all that hard to design a teeny daughter > board that fits into the 80386 socket, and carries an 80486. Is there a > real difference in design of the support circuitry, that an 80486 wouldn't > work in place of an 80386? From what I understand about the '486, to the > outside world (i.e., the motherboard), it shouldn't look much different from > an 80386. I would guess that the 80486 might have a couple extra control > lines that might need to be tied off somehow, but it should be able to start > up off of a standard 80386 BIOS, even. When '486 prices fall, this seems > like a potentially inexpensive upgrade - instead of getting a fast 80387, > for example. > > - David Hinds > dhinds@portia.stanford.edu I would have to agree. The design of the 486 compared to the 386 is not unlike the 386 is to the 286, for which there are many options for upgrades, only the most radical of which is motherboard replacement. I would have to beleive the boys at Intel are gonna be working overtime to get an Inboard/486 out on the market, once they get the bugs out of the 486, that is. Alex Cutrone v126gfl3@ubvms.cc.buffalo.edu
cb@sequoia.UUCP (Christopher D. Brown) (03/01/90)
?!?!?!? If you have a 386 why would you switch to a 486 instead of adding
a 387?
The only IMPORTANT differences, as I understand, are as follows:
1) Intel i$ $pending lot$ of dollar$ trying to make u$ believe
that the 486 is my$tically better than 386+387.
2) The 386+387 configuration is available and works.
What am I missing?
--
Christopher D. Brown
Digital: {uunet|texbell|cs.utexas.edu}!execu!cb
Analog: (512) 327-7070
Physical: Execucom, 108 Wild Basin Road, Two Wild Basin, Austin, TX 78764johnl@esegue.segue.boston.ma.us (John R. Levine) (03/02/90)
In article <9830@sequoia.UUCP> cb@sequoia.UUCP () writes: >?!?!?!? If you have a 386 why would you switch to a 486 instead of adding >a 387? At a given clock rate, a 486 is at least three times faster than a 386. While the 386 takes at least two and more often five or six cycles to execute an instruction, the 486 runs most of its instructions in one cycle each. The speedup for floating point is apparently even greater. When Intel went from the 286 to the 386, they added a lot of new function but the clocks per instruction remained about the same. The 486, on the other hand, is basically a performance upgrade with a tiny bit of new function to make it easier to implement multiprocessor systems. Then again, given that the 486 is still a member of the bug-of-the-month club, and that it costs considerably more, I'll wait. People who care about math speed should consider getting a Cyrix or IIT math chip. -- John R. Levine, Segue Software, POB 349, Cambridge MA 02238, +1 617 864 9650 johnl@esegue.segue.boston.ma.us, {ima|lotus|spdcc}!esegue!johnl "Now, we are all jelly doughnuts."
dhinds@portia.Stanford.EDU (David Hinds) (03/02/90)
In article <9830@sequoia.UUCP>, cb@sequoia.UUCP (Christopher D. Brown) writes: > ?!?!?!? If you have a 386 why would you switch to a 486 instead of adding > a 387? > > The only IMPORTANT differences, as I understand, are as follows: > 1) Intel i$ $pending lot$ of dollar$ trying to make u$ believe > that the 486 is my$tically better than 386+387. > 2) The 386+387 configuration is available and works. > > What am I missing? > It seems to me that the superiority of the 486 is a bit more than mystery. First, it sounds like it would be reasonable to expect a 486 stuck in a 386 board to run about 2X as fast as the 386 at the same clock speed and with the same support hardware, due to improved pipelining and the internal cache. Second, the 486 floating point unit is supposed to be like an order of magnitude faster than a 387. If adapting a 486 to a 386 socket is trivial, as it sounds like it might be, the cost of the upgrade would be little more than the cost of the 486. Sounds to me like an easy way to upgrade, no? A peripheral advantage of the 486 market is that it will put a lot of pressure on 386 and 387 prices, since a 386+387 needs to be quite a bit cheaper than a 486 for Intel to hope to save both markets. I don't know what a 486 costs now - as a guess, say $1000 for 25MHz? If this is cut in half over the next year, which would not be unreasonable, that would push similarly clocked 386+387 prices into the basement. -David Hinds dhinds@popserver.stanford.edu
wallwey@boulder.Colorado.EDU (WALLWEY DEAN WILLIAM) (03/02/90)
In article <241.25ec328b@ubvms.cc.buffalo.edu> v126gfl3@ubvms.cc.buffalo.edu writes: >............ The design of the 486 compared to the 386 is not unlike >the 386 is to the 286, for which there are many options for upgrades, only the >most radical of which is motherboard replacement. I would have to beleive the >boys at Intel are gonna be working overtime to get an Inboard/486 out on the >market, once they get the bugs out of the 486, that is. > >Alex Cutrone >v126gfl3@ubvms.cc.buffalo.edu If the recent past is an indication about "drop in proccessor boards", I would NOT consider getting or waiting for one. In the last year, it has really become as INEXPENSIVE to replace the MOTHER-BOARD as to use an "Inboard-type drop-in board". By actually replacing the mother-board you also forgo so many of the probablems associated with drop-in boards. Also mother-board replacements tend to yeild faster machines. To be Specific, when you use a "Drop in microproccessor board, you are still limited to using your original memory, and bus, not to memtion your system has more unneeded complexity. If you actually replace the mother-board, the whole system is actually deisgned for that microproccessor. To give an example, the i486 has some new special burst cache fill features. To use these features, the i486 has to have the memory system designed specially for it. Granted you can disable these features, but why. I do understand that if intel actually does make an Inboard/486 product, they will probably put the memory on the board. Even so, everything else will still be limited to the bus of the original mother board. Also trying to get all the memory on the Inboard might be a little hard with larger (lots of MB) systems. Finally, to change the mother board on most systems isn't that much harder than another card. Granted it will take maybe an hour and half, compared with just the half for drop in board, but I think the benifits are worth it!!! Just Some of My opinions Dean
cs4g6ag@maccs.dcss.mcmaster.ca (Stephen M. Dunn) (03/02/90)
In article <9830@sequoia.UUCP> cb@sequoia.UUCP () writes:
$?!?!?!? If you have a 386 why would you switch to a 486 instead of adding
$a 387?
Speed.
$The only IMPORTANT differences, as I understand, are as follows:
$ 1) Intel i$ $pending lot$ of dollar$ trying to make u$ believe
$ that the 486 is my$tically better than 386+387.
$ 2) The 386+387 configuration is available and works.
$What am I missing?
Well, I recall seeing a benchmark of a couple of 486/25 systems
against a couple of 386/33 systems in BYTE's February issue. Relative
to an 8 MHz AT, the 386 systems had a performance level of around 6,
while the 486 systems were at about 7 on a mix that didn't include
floating point. Floating point ran somewhat more than twice as fast
on the 486 machines as it did on the 386 machines.
Don't forget that a) Intel has made many instructions operate much
faster on the 486 than on the 386/387 (especially in floating point),
and b) all 486 systems have at least 8K of cache memory, and optionally
more, while 386 systems may not have any.
It isn't just mystic ... it's real.
--
Stephen M. Dunn cs4g6ag@maccs.dcss.mcmaster.ca
<std_disclaimer.h> = "\nI'm only an undergraduate!!!\n";
****************************************************************************
I Think I'm Going Bald - Caress of Steel, Rushalexew@watserv1.waterloo.edu (Alex E. Wielhouwer) (03/02/90)
In article <9673@portia.Stanford.EDU> dhinds@portia.Stanford.EDU (David Hinds) writes: > First, it sounds like it would be reasonable to expect a 486 >stuck in a 386 board to run about 2X as fast as the 386 at the same >clock speed and with the same support hardware, due to improved pipelining >and the internal cache. Second, the 486 floating point unit is supposed >to be like an order of magnitude faster than a 387. If adapting a 486 ----- Please correct me if I'm wrong, but in a multi-tasking environment, the floating point registers and stack need to be handled seperately on a '386/'387 system during task swaps, or, in some cases the '387 is ignored. (OS/2?). With a '486 environment, this should be much more efficient, i.e., faster. ============================================================================= Alex E. Wielhouwer, P.Eng., (519)885-1211 x3422 Department of Computing Services, University of Waterloo, 200 University Avenue W, Waterloo, Ontario, Canada. N2L 3G1 =============================================================================
dhinds@portia.Stanford.EDU (David Hinds) (03/03/90)
In article <1304@watserv1.waterloo.edu>, alexew@watserv1.waterloo.edu (Alex E. Wielhouwer) writes: > > Please correct me if I'm wrong, but in a multi-tasking environment, the > floating point registers and stack need to be handled seperately on a > '386/'387 system during task swaps, or, in some cases the '387 is ignored. > (OS/2?). With a '486 environment, this should be much more efficient, i.e., > faster. During task switches, the 386 saves its own context, but does not automatically save the floating point processor's context. It generates an interrupt at the first floating point operation following a task switch, so the system software can handle switching 387 contexts. As this feature was not present in the 80286, OS/2 probably doesn't know about it. OS/2 may do it manually whenever a task switch happens, or never do it - I don't know. The 386 method has the advantage of not unnecessarily saving and restoring the 387 context for tasks that don't use floating point. I don't know how the 486 does this, but one of the "new features to improve multitasking" may be to quickly save/restore the floating point context automatically. -David Hinds dhinds@popserver.stanford.edu
ralf@b.gp.cs.cmu.edu (Ralf Brown) (03/03/90)
In article <9729@portia.Stanford.EDU> dhinds@portia.Stanford.EDU (David Hinds) writes: } During task switches, the 386 saves its own context, but does not }automatically save the floating point processor's context. It generates }an interrupt at the first floating point operation following a task switch, }so the system software can handle switching 387 contexts. As this feature }was not present in the 80286, OS/2 probably doesn't know about it. OS/2 Certainly it was present in the 286. From the Intel 286 Programmer's Reference: The context of a processor extension (such as the 80287 numerics processor) is not changed by the task switch operation. ... The 80286 detects the first use of a processor extension after a task switch by causing the processor extension not-present exception (#7) if the TS bit is set. (page 11-5) -- {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.
davidsen@sixhub.UUCP (Wm E. Davidsen Jr) (03/03/90)
In article <9830@sequoia.UUCP> cb@sequoia.UUCP () writes: | The only IMPORTANT differences, as I understand, are as follows: | 1) Intel i$ $pending lot$ of dollar$ trying to make u$ believe | that the 486 is my$tically better than 386+387. | 2) The 386+387 configuration is available and works. Fools my benchmarks! I see the 486 about 2.2 times faster than a 386 at the same clock (I have seen 2.6 elsewhere, so that's close), and for f.p. the 486 is about 4 times faster than the 387. I have no idea where you got your information, but it seems to be completely wrong. -- bill davidsen - davidsen@sixhub.uucp (uunet!crdgw1!sixhub!davidsen) sysop *IX BBS and Public Access UNIX moderator of comp.binaries.ibm.pc "Getting old is bad, but it beats the hell out of the alternative" -anon
kdq@demott.COM (Kevin D. Quitt) (03/04/90)
On the other hand, (presumably) all of the bugs in the 386 are known, along with the work arounds. By the time we get to that point with the 486, Intel will be producing the 586. 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