coop4y44@bwdla28.bnr.ca (Takis Skagos) (04/04/91)
Hi. I've a question about the different chips in the 80386 family tree. I've heard the 386 refered to as 386SX, 386DX, and something else that I cannot remember. Does anybody out there know? Taki -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- BNR Ottawa Disclaimer: "I swear, they made me do it!" P. Takis Skagos -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
akcs.gregc@vpnet.chi.il.us (*Greg*) (04/04/91)
Takis Skagos scribbled: >Item: 2869 by coop4y44 at bwdla28.bnr.ca > Subj: Question about different 386's ... > Date: Thu Apr 04 1991 03:02 > > Hi. I've a question about the different chips in the 80386 family >tree. I've heard the 386 refered to as 386SX, 386DX, and something >else that I cannot remember. Does anybody out there know? Hi too.. The 80386 is very simple. 386SX is a 16 bit crunching chip. It processes 16 bits at a time while the 386DX is a 32 bit crunching chip. You are probably better off to get the DX. But for tight bugets; an SX is the way to go. ____ \GC/ Greg Clawson \/ Chicago IL. - The heart of America
c60b-1eq@e260-1f.berkeley.edu (Noam Mendelson) (04/05/91)
In article <27fb0c94-b35.1comp.sys.ibm.pc.hardware-1@vpnet.chi.il.us> akcs.gregc@vpnet.chi.il.us (*Greg*) writes: >Takis Skagos scribbled: >>Item: 2869 by coop4y44 at bwdla28.bnr.ca >> Subj: Question about different 386's ... >> Date: Thu Apr 04 1991 03:02 >> Hi. I've a question about the different chips in the 80386 family >>tree. I've heard the 386 refered to as 386SX, 386DX, and something >>else that I cannot remember. Does anybody out there know? >Hi too.. The 80386 is very simple. 386SX is a 16 bit crunching chip. It >processes 16 bits at a time while the 386DX is a 32 bit crunching chip. You >are probably better off to get the DX. But for tight bugets; an SX is the >way to go. I disagree. Even though the DX has 16 more pins, most of the time they're not being used. This is due to the fact that most MSDOS programs are 16-bit (with the exception of 386-specific programs). So unless you actually make use of all 32 pins, a 386SX is not a bad way to go. +==========================================================================+ | Noam Mendelson ..!agate!ucbvax!web!c60b-1eq | "I haven't lost my mind, | | c60b-1eq@web.Berkeley.EDU | it's backed up on tape | | University of California at Berkeley | somewhere." |
pschwart@vms.macc.wisc.edu (Paul Schwartz) (04/05/91)
In article <27fb0c94-b35.1comp.sys.ibm.pc.hardware-1@vpnet.chi.il.us> akcs.gregc@vpnet.chi.il.us (*Greg*) writes: >Takis Skagos scribbled: >>Item: 2869 by coop4y44 at bwdla28.bnr.ca >> Subj: Question about different 386's ... >> Date: Thu Apr 04 1991 03:02 >> Hi. I've a question about the different chips in the 80386 family >>tree. I've heard the 386 refered to as 386SX, 386DX, and something >>else that I cannot remember. Does anybody out there know? >Hi too.. The 80386 is very simple. 386SX is a 16 bit crunching chip. It >processes 16 bits at a time while the 386DX is a 32 bit crunching chip. Wrong, wrong, wrong. Data is transfered to and from the SX in 16-bit chunks, but its internal processing is equivalent to the DX, 32-bits. - Z - +-----------------------------------------------------------------------------+ | PauL M SchwartZ | There are many causes worth dying for, | | PSCHWART@macc.wisc.edu | but none worth killing for. | | PSCHWART@wiscmacc.BitNet | - Gandhi | +-----------------------------------------------------------------------------+
smsmith@hpuxa.acs.ohio-state.edu (Stephen M. Smith) (04/06/91)
c60b-1eq@e260-1f.berkeley.edu (Noam Mendelson) writes: akcs.gregc@vpnet.chi.il.us (*Greg*) writes: >>Takis Skagos scribbled: >>> Hi. I've a question about the different chips in the 80386 family >>>tree. I've heard the 386 refered to as 386SX, 386DX, and something >>>else that I cannot remember. Does anybody out there know? >>Hi too.. The 80386 is very simple. 386SX is a 16 bit crunching chip. It >>processes 16 bits at a time while the 386DX is a 32 bit crunching chip. You >>are probably better off to get the DX. But for tight bugets; an SX is the >>way to go. > >I disagree. Even though the DX has 16 more pins, most of the time they're >not being used. This is due to the fact that most MSDOS programs are 16-bit >(with the exception of 386-specific programs). So unless you actually make >use of all 32 pins, a 386SX is not a bad way to go. Hmmm..... Maybe a bit of clarification is needed: SX and DX chips are both 32-bit processing chips. BOTH will run 386-specific programs, DOS or otherwise. The number of external pins on the SX does not affect its being a true 386 processor, nor do they make it incompatible with 32-bit "protected" mode software. The difference in the two chips is that the DX can access anything EXTERNAL to the chip via a true 32-bit channel, whereas the SX accesses external things via a 16-bit channel. This includes memory, peripherals, or whatever else. Also, most MSDOS programs are not 16-bit. They are 8-bit. How else do you think I've been running them on my XT?????? Steve Smith
kaleb@thyme.jpl.nasa.gov (Kaleb Keithley) (04/06/91)
In article <1991Apr5.213536.12925@magnus.acs.ohio-state.edu> smsmith@hpuxa.acs.ohio-state.edu (Stephen M. Smith) writes: >c60b-1eq@e260-1f.berkeley.edu (Noam Mendelson) writes: >>akcs.gregc@vpnet.chi.il.us (*Greg*) writes: >>>Takis Skagos wrote: >>>> Hi. I've a question about the different chips in the 80386 family >>>Hi too.. The 80386 is very simple. 386SX is a 16 bit crunching chip. It >>I disagree. Even though the DX has 16 more pins, most of the time they're >Hmmm..... Maybe a bit of clarification is needed: >SX and DX chips are both 32-bit processing chips. BOTH will Let's sing the 386 song. (to the tune of Revolution #9, by the Beatles) A 386, is a 386, is a 386, is.... you get the idea. The Intel 386 family is a 32 bit processor. Variants known as SX and DX exist. The SX has a 16-bit data bus. the DX has a 32-bit data bus. The DX has a longer instruction pre-fetch queue. In general the DX is a) more expensive, and b) more powerfull. >Also, most MSDOS programs are not 16-bit. They are 8-bit. How >else do you think I've been running them on my XT?????? Close, but no cigar. The 8086/8088 (and 80186/80188/80286) are *all* 16-bit CPUs. The 8086/80186/80286 have 16-bit data busses, while the 8088/80188 have 8-bit data busses. Too bad Intel didn't follow their own convention and call the 386SX an 80388 :-) BTW, all MSDOS programs *are* 16-bit programs because they run in 16-bit (and 32-bit) CPUs. -- Kaleb Keithley kaleb@thyme.jpl.nasa.gov Meep Meep Roadrunner Veep veep Quayle
phil@cs.mcgill.ca (Philip LOCONG) (04/06/91)
>[debate about usefullness of 386DX vs 386SX with one (con)argument > being that most MSDOS programs are 16bit] >Also, most MSDOS programs are not 16-bit. They are 8-bit. How >else do you think I've been running them on my XT?????? Doesn't an XT have a 8086 ? (which is 16-bit in the same way 386SX is 32-bit) :-) (However, I do agree that most or "very very many" programs are 8-bit)
torvalds@cc.helsinki.fi (04/06/91)
In article <1991Apr5.023634.549@agate.berkeley.edu>, c60b-1eq@e260-1f.berkeley.edu (Noam Mendelson) writes: >> [ discussion on 386DX vs 386SX deleted ] > > I disagree. Even though the DX has 16 more pins, most of the time they're > not being used. This is due to the fact that most MSDOS programs are 16-bit > (with the exception of 386-specific programs). So unless you actually make > use of all 32 pins, a 386SX is not a bad way to go. I'm not disagreeing outright, but I do have one point: dos (and all 16-bit programs) won't actually use the 16 high bits for data transfer (the pins are used though), but instruction prefetching goes on as before. I don't have statistics or anything, but I's say MUCH more than 50% of all memory fetches are instruction prefetching. Getting 4 bytes instead of 2 at a time into the prefetch queue should make a notable difference. Anybody have any actual timing differences? The point is that the 16 extra pins should be a big win even if not used explicitly by the programmer. Linus Torvalds torvalds@cc.helsinki.fi PS. Speed shouldn't really be an issue here. As far as I know the SX is made in max 20MHz versions, most (?) DX's are >20MHz. The big difference between the DX and SX is in price - if you can live with a slower system, by all means go for the SX.
smsmith@hpuxa.acs.ohio-state.edu (Stephen M. Smith) (04/06/91)
kaleb@thyme.jpl.nasa.gov (Kaleb Keithley) writes: >smsmith@hpuxa.acs.ohio-state.edu (Stephen M. Smith) writes: >>Also, most MSDOS programs are not 16-bit. They are 8-bit. How >>else do you think I've been running them on my XT?????? > >Close, but no cigar. The 8086/8088 (and 80186/80188/80286) are *all* >16-bit CPUs. The 8086/80186/80286 have 16-bit data busses, while the >8088/80188 have 8-bit data busses. Too bad Intel didn't follow their >own convention and call the 386SX an 80388 :-) > >BTW, all MSDOS programs *are* 16-bit programs because they run in 16-bit >(and 32-bit) CPUs. Yes, that's correct. Oops. Stephen Smith
c60b-1eq@e260-1d.berkeley.edu (Noam Mendelson) (04/06/91)
In article <1991Apr5.213536.12925@magnus.acs.ohio-state.edu> smsmith@hpuxa.acs.ohio-state.edu (Stephen M. Smith) writes: >c60b-1eq@e260-1f.berkeley.edu (Noam Mendelson) writes: >akcs.gregc@vpnet.chi.il.us (*Greg*) writes: >>>Takis Skagos scribbled: >>>> Hi. I've a question about the different chips in the 80386 family >>>>tree. I've heard the 386 refered to as 386SX, 386DX, and something >>>>else that I cannot remember. Does anybody out there know? >>>Hi too.. The 80386 is very simple. 386SX is a 16 bit crunching chip. It >>>processes 16 bits at a time while the 386DX is a 32 bit crunching chip. You >>>are probably better off to get the DX. But for tight bugets; an SX is the >>>way to go. >>I disagree. Even though the DX has 16 more pins, most of the time they're >>not being used. This is due to the fact that most MSDOS programs are 16-bit >>(with the exception of 386-specific programs). So unless you actually make >>use of all 32 pins, a 386SX is not a bad way to go. >Hmmm..... Maybe a bit of clarification is needed: >SX and DX chips are both 32-bit processing chips. BOTH will >run 386-specific programs, DOS or otherwise. The number of >external pins on the SX does not affect its being a true >386 processor, nor do they make it incompatible with >32-bit "protected" mode software. >The difference in the two chips is that the DX can access anything >EXTERNAL to the chip via a true 32-bit channel, whereas the SX >accesses external things via a 16-bit channel. This includes >memory, peripherals, or whatever else. True, but I'm not arguing that. I was simply pointing out the differences between them. >Also, most MSDOS programs are not 16-bit. They are 8-bit. How >else do you think I've been running them on my XT?????? No, MSDOS programs are 16-bit. If your XT is equipped with an 8088 (instead of an 8086), then you technically have an 8-bit processor; however, the situation is similar to that of a 386SX. The bus is 8 bits wide, but the processor can internally handle 16 bits. +==========================================================================+ | Noam Mendelson ..!agate!ucbvax!web!c60b-1eq | "I haven't lost my mind, | | c60b-1eq@web.Berkeley.EDU | it's backed up on tape | | University of California at Berkeley | somewhere." |
lbr@holos0.uucp (Len Reed) (04/08/91)
>>Also, most MSDOS programs are not 16-bit. They are 8-bit. How >>else do you think I've been running them on my XT?????? It's amazing that someone could correctly describe the 80386SX as internally 32-bit and externally (i.e., data path) 16 bit and then completly miss the fact that the 8088 is the same animal in the 16/8 bit world. DOS runs in 16-bit mode; full-word (16-bit) accesses take 2 accesses on an 8088 and 1 on 8086, 80186 (not comonly used in PCs), 80286 and above. >Doesn't an XT have a 8086 ? >(which is 16-bit in the same way 386SX is 32-bit) No, it has an 8088 which is 16-bit in the same way the 386SX is 32-bit. An 8086 is a full 16-bit processor. 8086: 16/16 8088: 16/8 80186: 16/16, additional instructions (e.g., PUSHA), on-board peripheral control, timer, and DMA, faster than 8086 at same clock speed. Often used in embedded systems. 80286: 16/16, has 80186 extra instructions. Has "protected" mode for multitasking operating systems (e.g., Xenix). Can address 16 Mbytes in protected mode. 80386: (also called 80386DX). 32/32. 80386SX: 32/16. Fully software compatible with DX, but will run slower especially in 32-bit mode (e.g., running Unix). 80486: Intel refers to this as i486, which is a trademark. Sort of like a 80386DX and a 80387 math coprocessor on one chip. Also has 8Kbyte cache memory and a few extra instructions. -- Len Reed Holos Software, Inc. Voice: (404) 496-1358 UUCP: ...!gatech!holos0!lbr
ntaib@silver.ucs.indiana.edu (Nur Iskandar Taib) (04/09/91)
>Close, but no cigar. The 8086/8088 (and 80186/80188/80286) are *all* >16-bit CPUs. The 8086/80186/80286 have 16-bit data busses, while the >8088/80188 have 8-bit data busses. Too bad Intel didn't follow their >own convention and call the 386SX an 80388 :-) Or the 8088 the 8086SX.. hehehehehe -- ------------------------------------------------------------------------------- Iskandar Taib | The only thing worse than Peach ala Internet: NTAIB@AQUA.UCS.INDIANA.EDU | Frog is Frog ala Peach Bitnet: NTAIB@IUBACS ! -------------------------------------------------------------------------------
ntaib@silver.ucs.indiana.edu (Nur Iskandar Taib) (04/09/91)
>> I disagree. Even though the DX has 16 more pins, most of the time they're >> not being used. This is due to the fact that most MSDOS programs are 16-bit >> (with the exception of 386-specific programs). So unless you actually make >> use of all 32 pins, a 386SX is not a bad way to go. So how do you exceed 16 MByte RAM if you're only using 16 bit addresses? 8-) I think the SX actually has over 100 pins 8-) -- ------------------------------------------------------------------------------- Iskandar Taib | The only thing worse than Peach ala Internet: NTAIB@AQUA.UCS.INDIANA.EDU | Frog is Frog ala Peach Bitnet: NTAIB@IUBACS ! -------------------------------------------------------------------------------
kls30@duts.ccc.amdahl.com (Kent L Shephard) (04/11/91)
In article <1991Apr5.213536.12925@magnus.acs.ohio-state.edu> smsmith@hpuxa.acs.ohio-state.edu (Stephen M. Smith) writes: >The difference in the two chips is that the DX can access anything >EXTERNAL to the chip via a true 32-bit channel, whereas the SX >accesses external things via a 16-bit channel. This includes >memory, peripherals, or whatever else. > >Also, most MSDOS programs are not 16-bit. They are 8-bit. How >else do you think I've been running them on my XT?????? Because the 8088 is to a 8086 as a 80386sx is to a 80386dx. The 8088 IS 16bit it just accesses memory 8bits at a time. > >Steve Smith -- /* -The opinions expressed are my own, not my employers. */ /* For I can only express my own opinions. */ /* */ /* Kent L. Shephard : email - kls30@DUTS.ccc.amdahl.com */
jerry@polygen.uucp (Jerry Shekhel) (04/13/91)
kaleb@thyme.jpl.nasa.gov (Kaleb Keithley) writes: > >Close, but no cigar. The 8086/8088 (and 80186/80188/80286) are *all* >16-bit CPUs. The 8086/80186/80286 have 16-bit data busses, while the >8088/80188 have 8-bit data busses. Too bad Intel didn't follow their >own convention and call the 386SX an 80388 :-) > Actually, since the 386DX is a 32 bit processor, it should have been named the 80382 (the 2 standing for 32), and the 386SX should have been the 80386. Now they're coming out with the 486SX, which is 32-bit, but without on-board coprocessor. That makes it all the more confusing, since SX no longer implies a 16-bit data path, but rather a somewhat less powerful version of the non-SX. Maybe it just stands for "SuX" :-) > >-- >Kaleb Keithley kaleb@thyme.jpl.nasa.gov > -- +-------------------+----------------------+---------------------------------+ | JERRY J. SHEKHEL | POLYGEN CORPORATION | When I was young, I had to walk | | Drummers do it... | Waltham, MA USA | to school and back every day -- | | ... In rhythm! | (617) 890-2175 | 20 miles, uphill both ways. | +-------------------+----------------------+---------------------------------+ | ...! [ princeton mit-eddie bu sunne ] !polygen!jerry | | jerry@polygen.com | +----------------------------------------------------------------------------+