Re: Why I've Dropped In

On 6/13/2025 11:11 AM, Thomas Koenig wrote:

Stephen Fuld <sfuld@alumni.cmu.edu.invalid> schrieb:
 

Even with that provision, it would not have worked.

>
All I can say is that it worked in several other contemporaneous
architectures.  Scott gave one example, the Burroughs Medium systems.
The Univac 1108 and followons is another.  There may be others, perhaps
the some of CDC systems as they and the Univac systems shared a common
designer (Seymour Cray).

 The problem of the /360 was that they put their base registers in
user space.

Yes.  That is the point I was making.  Now I am lost as to why you said above "even with that provision, it would not have worked."

The other machines made it invisible from user space
and added its contents to every memory access.  This does not take
up opcode space and allows swapping in and out of the whole process,
which was a much better solution for the early 1960s.

You and I are in violent agreement!  Only John seems to disagree.

(Actually, I
believe the UNIVAC had two, one for program and one for data).

Correct.  I didn't want to complicate the discussion.  The separation of instructions from data allowed the OS to put them in different memory modules, thus allowing simultaneous access to the current instruction's operand and to the next instruction fetch, thus dramatically improving performance.  (Remember, no cache).  And later follow on systems such as the 1110 actually had two sets of Instruction and Data bases, and later machines went to 16 base registers.  This is similar to the progression Scott talked about for the Burroughs medium systems.
--
  - Stephen Fuld
(e-mail address disguised to prevent spam)