Re: Future architectures

Bernd Paysan <bernd.paysan@xxxxxx> writes:
Joe Pfeiffer wrote:
Comes from the Harvard Mark I, which used a paper tape for instructions
and some sort of electromechanical storage for data.

That goes back to the first Zuse machine. The actual advance was von
Neumann's, because he found that instructions are just another form of
data. However, many computer languages haven't caught up to that idea, and
still only manipulate data, but don't provide meta-programming features.

I'm missing the connection between our posts -- I'm not familiar with
the details of the Zuse machines (one of those things I keep meaning
to learn about...), but what confuses me is how von Neumann's advance
was related to the Harvard architectures -- after all, in the Harvard
architectures you *couldn't* manipulate instructions like data.

FWIW, when I was an undergrad at the University of Washington, we were
taught that the same-memory and different-memory options were called
"Princeton" (after the location of the IAS) and "Harvard", and they
were both von Neumann machines. I'm also used to split-cache being
recycled and called Harvard, while unified cache being called
Princeton.

I think the wikipedia entry Nick was referring to may be the first
time I've seen Harvard contrasted with von Neumann (of course, that
could easily mean I've missed a lot).
.