• Please review our updated Terms and Rules here

Has there ever been 68000-bit processors?

W

Windows XP

Member
Joined
Jul 5, 2019
Messages
11
I am :confused: that old Macs have 68k processors, but there is no such thing as 68000-bit, the current Macs are 64-bit! What am I getting wrong here?
 
The 68k refers to a Motorola part number. 64 bits refers to the native data bus width of the processor.

The 68000 CPU was a Motorola 16 bit processor with an external 16 bit data bus. The 68008 CPU was a Motorola 16 bit processor with an external 8 bit data bus. Internal logic read two sequential 8 bit data bytes to form a 16 bit word internally.

A modern iMac contains a 64 bit Intel processor.

Hope that clears up the confusion?

Dave
 
Al Kossow said:
is the 68020 a 16 bit processor?
Click to expand...

I'm pretty sure people have come to blows (beer bottles smashed on the bar, the whole enchilada) over whether or not the 68000 should be called a 16 or 32 bit processor, but I think the 32-bit-ness is pretty unequivocal in the case of the 68020.
 
Yes...

Is a Zilog Z80 an 8 or 16 bit machine? It has 8 bit registers internally, but they can be ‘joined together’ to form 16 bit register pairs. To further confuse things, you can perform 16-bit arithmetic on some of the register pairs. I would still call the Z80 an 8-bit processor however.

Churches have been built on less...

Dave
 
I generally go by the width of the ALU. Not always, however. There are some serial ALU minis and microprocessors; it just wouldn't be right to call them "one bit". :)
 
Well, you could classify by the largest integer on which the complete range of arithmetic and logical operations could be performed. That would confine the Z80 to 8 bits (the set of 16-bit operations is incomplete) and the 68K to 32 bits; the 8086 would be 16 bits.

This, of course, doesn't count SIMD (vector) architectures.

Using the width of the memory bus doesn't work either. Consider the CDC STAR--the memory interface to the CPU is 528 bits (includes parity). But it's nominally a 64 bit architecture. The big bus is for vector processing.

Then there are the old variable-length systems (e.g. 1401, 1620). If you used the "complete set of arithmetic operations" criterion, even the lowly 360/20 would turn in some very high numbers (most of the arithmetic is implemented as SS packed decimal).

It is a puzzlement.
 
Eudimorphodon said:
I'm pretty sure people have come to blows (beer bottles smashed on the bar, the whole enchilada) over whether or not the 68000 should be called a 16 or 32 bit processor, but I think the 32-bit-ness is pretty unequivocal in the case of the 68020.
Click to expand...

I've obviously been going to the wrong bars all my life.
daver2 said:
Yes...

Is a Zilog Z80 an 8 or 16 bit machine? It has 8 bit registers internally, but they can be ‘joined together’ to form 16 bit register pairs. To further confuse things, you can perform 16-bit arithmetic on some of the register pairs. I would still call the Z80 an 8-bit processor however.

Churches have been built on less...

Dave
Click to expand...

I think that was either a programmer's trick for dealing with 16-bit arithmetic or a kind of hackish bit-slice technique used on some systems that tried to use them in tandem to create a 16-bit or greater system, depending on what you're talking about. There was a lot of hacky stuff going on back then with the price difference between consumer grade chips and the more expensive 16+ bit chips.
 
If you should want to play with a real 68000 and not one of the follow on's. You should get a Canon Cat. Although, you need to learn a little Forth, is can have an assembler to work at the machine level. You can use the Cat's exculent editor for your source code and then compile it on the CAT.
Do note that if you get a machine that seems to have a working disk drive, do not attempt to remove the floppy until you talk to me. The drives are no longer available and can be easily destroyed.
Dwight
 
You must log in or register to reply here.
Back
Top