Incorrect. The 8086 and 8088 are both 16 bit processors capable of running the same binaries. The architecture and instruction sets of both chips are identical.
Identical? If having 40 pins means "architecture" to you; there are issues. The first one would be "...find the pin
D8 on the 8088"
If you're unaware of the differences between instructions of the 88 and 86, I'm guessing you've never programmed hardware support. (The 88 requires Min/Max and High/Low byte registers to determine which half of the 16-bit value it's attempting to emulate... the 86 has none of this. Any ASM experience from back then, by chance? I would have imagined you to know this.)
The 8088 has an 8 bit data bus so it needs to do 2 bus transactions for each 16 bit transfer. So it's throughput is slower than an 8086 running at the same clock speed. It also had a 4 byte prefetch queue.
You just said it yourself... 8 bit bus... versus the 86's 16 bit bus
"The architecture and instruction sets of both chips are identical."
If you believe all of the above to be true, here's a task for you:
- Take either release of the 8088 and socket it onto a generic breadboard w/ shielding
- I'll give you a basic schematic to follow to add a crystal and filters to match
- Find a way to pump 16 "ones and zeros" out of eight pins with no external hardware added
In short, without the added hardware normally found on 8088 dedicated
motherboards, the processor itself has no way to communicate with the outside world in 16 bit words.
For all of us hardware junkies who were forced to study tube theory in the transistor age, there was one common with logic devices: The number of pins (or tubes!) available to read or represent data is the maximum number of bits the
hardware is capable of. I couldn't care less about emulation (I'll explain.)
The 80286 is a (true) 16 bit microprocessor (doesn't matter what instruction set / commands / etc. it can speak or be spoken to - - it can pump 16 "ones and zeros" in a single Machine Cycle via RAM or I/O.)
The 80386 is a (true) 16 bit microprocessor -- again, doesn't matter that it "emulates" 32 bit instructions -- it can only transmit 16 "bits and bobs" in a single breath.
Emulation, such as the 8088 or even the 386 have one thing in common... due to their fake nature, every shift they have to do in order to reach their intended number of data bits costs Machine Cycles; in turn, cutting speed (Hz) in
half each time. An 8088 running under a 5mhz crystal will operate instructions at a little under 2.5Mhz. This "expense" was perfectly acceptable for shops building embedded devices and the likes -- as SPEED was less important than cost and addressing.
For example, there's NOTHING stopping anyone from taking an 8088 and re-flashing it to cycle FOUR times (instead of two) in turn, creating an array of 32 bits of data. Back in the day of 8088, the S100 or ISA BUS was never intended to carry 32 "ones and zeros" so nobody bothered doing it (that I'm AWARE of... never know!!!) Bare in mind, anything over 8 bits from the 88 has to be managed outside of the processor and the CPU can only assume success without an elaborate self-checking mechanism that would burn even more cycles.
If your definition of "Processor" means merely something that processes information, then I'll concede over language barriers. In this case, it's a 16 bit
information processor
regardless of how it gets there.
If your definition of "Processor" comes from "Microprocessor" which in turn defines the Integrated Circuit (Chip) which an MPU is defined by hardware through the number of address and data lines it has available and the speed in which it can process those lines in "ticks"... the 88's (intentionally) an 8-bit microprocessor capable of emulating 16 bit instruction when governed and assisted by external hardware allowing it to fulfill the needs of the E/ISA bus and the extended RAM BUS. (...OR, when NOT being implemented as a motherboard cake topper, you had the option to utilize the 88 as a full-fledged 8 bit micro OR, if you wanted to supply the necessary hardware, you could do the same 16bit emulation -- which wasn't all that common outside of control circuits where more than 250 logic ladders were required and they wanted everything under one roof.)
LOL1 - in the end... it was a lot easier to teach using a Z80 as opposed to i88 in 8-bit.
LOL2 - Tandy initially thought so, too!
(...someone in the audience mumbles... "who's Tandy?" Someone else whimpers "...I still wanna know what an 8088 is!?")
LOL3 - I had a client who just recently passed at age 97 who still had the 8086/87 system we built and programmed back in the 80s which we later upgraded the mobo to a 286 running our custom DOS foundry app for repair shop logs and diagnostics and a program called
PFS: First Choice for invoicing and inventory which it was never really intended for but was molded to do so! Decades went by...
Both apps were still alive and well as they maintained a 60+ year old TV/Electronics Repair Shop in N.E. Ohio years later sitting next to a newer box running XP and Quick Books.
...invoices still being printed on a dot-matrix "screamer" - I didn't know they even MADE continuous form paper anymore!!
Welcome
FAQ
Downloads
Wiki
