Re: Is microprocessor an integrated circuit???

From: keith (krw_at_att.bizzzz)
Date: 01/29/05 

Date: Sat, 29 Jan 2005 12:02:19 -0500

On Sat, 29 Jan 2005 11:12:05 +0000, Andrew Holme wrote:

> Keith Williams wrote:
>> In article <pan.2005.01.28.20.31.32.28264@example.net>,
>> richgrise@example.net says...
>>> On Fri, 28 Jan 2005 13:01:11 -0500, Keith Williams wrote:
>>>
>>>> In article <ctdto9$52o$1@eskinews.eskimo.com>, mzenier@eskimo.com
>>>> says...
>>>>> In article <uZTJd.187$Eh5.115@trnddc04>, Bradley1234
>>>>> <someone@yahoo.com> wrote:
>>>>>> PLEASE show an example of a microprocessor that doesnt use
>>>>>> microcode
>>>>>
>>>>> Most of them have their control unit logic implemented in a PLA
>>>>> (Programmable Logic Array) which directly implements a two level
>>>>> logic equation. With microcode, there would be an address that
>>>>> was decoded
>>>>> to provide a word (or row) of the ROM's contents. There's no such
>>>>> thing
>>>>> as an address in a PLA, just inputs and outputs.
>>>>
>>>> One can also argue (rightly) that a PLA is nothing but a sparse ROM,
>>>> thus microcode.
>>>
>>> I'd say, the definitive factor here is, can you write a program? Is
>>> there an instruction set?
>>
>> A microprocessor without an instruction set is, umm, useless?
>
> My preferred criteria is : Is there a micro-program counter?

Ah, so what about a processor that has no program counter? Is that not a
processor? Can it not be programmed? Does it not have programs?

> On reflection: this makes the 6502 microcoded because the instruction
> register and T state counter, which feed into the logic array, can be
> thought of as high and low words of a micro-program address.
>
>> I can list this "PLA" microcode (I have the power;), but it never
>> exists in hardware. It's converted by the synthesis tools from VHDL
>> into gates. What does that make it? Is it microprogrammed? The
>> source sure looks like a PLA. It's even called a PLA in the source.
>> It's random logic on the chip though.
>
> Is there anything in your processor that resembles a micro-program
> counter?

There isn't even a program counter, at least one that exists as an entity
that can be identified as such. It's not a little strange processor off
in the corner of its own little world. It's fairly mainstream. These
concepts of "microprogrammed or hard-wired" are rather old and the
difference is really blury these days.

>> I'm telling you that "thar be dragons" if you insist on defining things
>> with nice black lines.
>
> Of course; but sometimes simplication / generalisation can be useful /
> helpful for getting ideas across.

For education, sure. Most of the distinctions we're talking about aren't
really so much of a distinction at all. FPGAs, for example, use the same
SRAMs as "gates" for doing logic operations, registers, and memory. 

-- 
  Keith
Quantcast