By: rwessel (robertwessel.delete@this.yahoo.com), April 10, 2017 6:37 am
Room: Moderated Discussions
wumpus (lost.delete@this.in-a.cave.net) on April 10, 2017 6:31 am wrote:
> rwessel (robertwessel.delete@this.yahoo.com) on April 9, 2017 11:05 pm wrote:
> > Ireland (boh.delete@this.outlook.ie) on April 9, 2017 12:17 pm wrote:
> > > Before the personal computer Intel thing, really kicked into another gear (Linus was able
> > > to relate to this - mentioning that, the IBM personal computer, got built around an 8080 too, as opposed
> > > to an 8086 microprocessor, and for the reasons that he explained in one of the comments below).
> >
> >
> > The PC was built around an 8088, not anything in the 8080 family. IBM did build some earlier systems
> > around the 8085 (notable the S/23 Datamaster), but it had little to do with the eventual IBM PC. My
> > understanding of some of the histories I've read is that IBM at least considered an 8080 (presumably
> > actually 8085 or Z-80, given the time frame) based system, but that idea was rejected quite early on.
>
> My understanding was that much of the confusion could come about thanks to plenty of the support chips
> already were in use for the 8085/8080*. Since IBM pretty much copied the Intel app notes, what you
> wound up with was mostly Intel chips. This gave you nearly an 8086 at much lower prices.
It wasn't hard to get the 8086 to do byte addressed memory as well, and so could have largely used the same support chips as the 8088. In many cases those *were* the 8080 devices, Intel had taken of care to make the busses fairly compatible. In some cases there were newer device, better suited to the 8086/x. In other's, IBM didn't even use the best available 8080 device (using the older 8237 DMA controller instead of the newer 8257).
Probably the crucial thing would have been the cost of doubling of the width of memory, ROM and the I/O in order to maintain decent performance.
> * I'd assume they were 8085 parts, based on voltages needed. Did the 8085 have bit addressability?
ISA-wise, the 8085 was an 8080 with slightly different timing, and two new instructions (which were not frequently used). Neither had any bit operations, except the usual shifts and logical instructions.
> I remember
> using an intel assembler on a 8080-variant and have been pretty sure it was an 8085 (well a Phillips microcontroller
> based on said architecture). My memory doesn't seem to line up with the 8051 at all.
The 8051 is an embedded controller, and pretty much completely unrelated to the 8080 or 8086, and really unsuitable to building a "real" computer.
> rwessel (robertwessel.delete@this.yahoo.com) on April 9, 2017 11:05 pm wrote:
> > Ireland (boh.delete@this.outlook.ie) on April 9, 2017 12:17 pm wrote:
> > > Before the personal computer Intel thing, really kicked into another gear (Linus was able
> > > to relate to this - mentioning that, the IBM personal computer, got built around an 8080 too, as opposed
> > > to an 8086 microprocessor, and for the reasons that he explained in one of the comments below).
> >
> >
> > The PC was built around an 8088, not anything in the 8080 family. IBM did build some earlier systems
> > around the 8085 (notable the S/23 Datamaster), but it had little to do with the eventual IBM PC. My
> > understanding of some of the histories I've read is that IBM at least considered an 8080 (presumably
> > actually 8085 or Z-80, given the time frame) based system, but that idea was rejected quite early on.
>
> My understanding was that much of the confusion could come about thanks to plenty of the support chips
> already were in use for the 8085/8080*. Since IBM pretty much copied the Intel app notes, what you
> wound up with was mostly Intel chips. This gave you nearly an 8086 at much lower prices.
It wasn't hard to get the 8086 to do byte addressed memory as well, and so could have largely used the same support chips as the 8088. In many cases those *were* the 8080 devices, Intel had taken of care to make the busses fairly compatible. In some cases there were newer device, better suited to the 8086/x. In other's, IBM didn't even use the best available 8080 device (using the older 8237 DMA controller instead of the newer 8257).
Probably the crucial thing would have been the cost of doubling of the width of memory, ROM and the I/O in order to maintain decent performance.
> * I'd assume they were 8085 parts, based on voltages needed. Did the 8085 have bit addressability?
ISA-wise, the 8085 was an 8080 with slightly different timing, and two new instructions (which were not frequently used). Neither had any bit operations, except the usual shifts and logical instructions.
> I remember
> using an intel assembler on a 8080-variant and have been pretty sure it was an 8085 (well a Phillips microcontroller
> based on said architecture). My memory doesn't seem to line up with the 8051 at all.
The 8051 is an embedded controller, and pretty much completely unrelated to the 8080 or 8086, and really unsuitable to building a "real" computer.
