Prev: Check out POASM
Next: Blocks and compilers
From: Jonathan Bartlett on 3 Jan 2006 10:12
I am not much on computer history, but apparently someone who read my
book took exception to the idea that von Neumann developed the basic
architecture of the computer.

Here's what I said:

"Modern computer architecture is based off of an architecture called the
Von Neumann architecture, named after its creator. The Von Neumann
architecture divides the computer up into two main parts - the CPU (for
Central Processing Unit) and the memory. This architecture is used in
all modern computers, including personal computers, supercomputers,
mainframes, and even cell phones."

Here's what the writer said:

"What the heck, has the cult-leader and sophist John von Neumann to do,
with our modern computer's architecture? This architecture what you are
talking about was invented by Konrad Zuse in 1936, a few decades before
von Neumann got the nobel-price for cripping from Zuse. But actually
also Zuse was less a genius like von Neumann. The real discoveries our
today's computers are based on, comes from Kepler, Schikard, Leibniz and
Babbage."

To what extent is he right or wrong? Did von Neumann's ideas originate
with Zuse? Of course these are all based on past ideas, but did the
separation of computation and memory, and the program/data
single-storage paradigm originate before either Zuse or von Neumann? I
was hoping someone better versed in computer history could give me some
clues.

Jon
----
Learn to program using Linux assembly language
http://www.cafeshops.com/bartlettpublish.8640017
From: Alex McDonald on 3 Jan 2006 11:59
Jonathan Bartlett wrote:
> I am not much on computer history, but apparently someone who read my
> book took exception to the idea that von Neumann developed the basic
> architecture of the computer.
>
> Here's what I said:
>
> "Modern computer architecture is based off of an architecture called the
> Von Neumann architecture, named after its creator. The Von Neumann
> architecture divides the computer up into two main parts - the CPU (for
> Central Processing Unit) and the memory. This architecture is used in
> all modern computers, including personal computers, supercomputers,
> mainframes, and even cell phones."
>
> Here's what the writer said:
>
> "What the heck, has the cult-leader and sophist John von Neumann to do,
> with our modern computer's architecture? This architecture what you are
> talking about was invented by Konrad Zuse in 1936, a few decades before
> von Neumann got the nobel-price for cripping from Zuse. But actually
> also Zuse was less a genius like von Neumann. The real discoveries our
> today's computers are based on, comes from Kepler, Schikard, Leibniz and
> Babbage."
>
> To what extent is he right or wrong? Did von Neumann's ideas originate
> with Zuse? Of course these are all based on past ideas, but did the
> separation of computation and memory, and the program/data
> single-storage paradigm originate before either Zuse or von Neumann? I
> was hoping someone better versed in computer history could give me some
> clues.
>
> Jon
> ----
> Learn to program using Linux assembly language
> http://www.cafeshops.com/bartlettpublish.8640017

von Neumann architecture refers to the treatment of data and code as the
same thing; the CPU interprets "data". One of the alternatives is the
Harvard architecture, where code and data are separated.

Your critic's reference to "cult-leader and sophist" seem wide of the
mark; perhaps they're anti-Semitic comments, as he was Jewish. His Nobel
was in economics, not "a nobel-price for cripping from Zuse".

Leibniz discovered binary; Kepler I can't see the connection with at
all. Babbage was more an engineer than a scientist, and his "analytical
calculator" didn't work until well after his death (1991 in fact).
Wilhelm Schickard's contribution was a calculator too; it isn't
considered a computer, as it wasn't programmable, and his design seems
to have influenced no-one else.

Zuse is considered to have created the first working programmable
computer. The rest is interpretation of history.

--
Regards
Alex McDonald
From: randyhyde@earthlink.net on 3 Jan 2006 13:13

Jonathan Bartlett wrote:
> I am not much on computer history, but apparently someone who read my
> book took exception to the idea that von Neumann developed the basic
> architecture of the computer.
>
> Here's what I said:
>
> "Modern computer architecture is based off of an architecture called the
> Von Neumann architecture, named after its creator. The Von Neumann
> architecture divides the computer up into two main parts - the CPU (for
> Central Processing Unit) and the memory. This architecture is used in
> all modern computers, including personal computers, supercomputers,
> mainframes, and even cell phones."
>
> Here's what the writer said:
>
> "What the heck, has the cult-leader and sophist John von Neumann to do,
> with our modern computer's architecture? This architecture what you are
> talking about was invented by Konrad Zuse in 1936, a few decades before
> von Neumann got the nobel-price for cripping from Zuse. But actually
> also Zuse was less a genius like von Neumann. The real discoveries our
> today's computers are based on, comes from Kepler, Schikard, Leibniz and
> Babbage."
>
> To what extent is he right or wrong? Did von Neumann's ideas originate
> with Zuse? Of course these are all based on past ideas, but did the
> separation of computation and memory, and the program/data
> single-storage paradigm originate before either Zuse or von Neumann? I
> was hoping someone better versed in computer history could give me some
> clues.
>


Hi Jonathon,
Today, most people recognize the fact that Von Neumann got his ideas
for a shared-memory computer from some peers that he was working with
at the time (John William Mauchly and J. Presper Eckert?). There is an
article about Zuse vs. VN here:
http://irb.cs.tu-berlin.de/~zuse/Konrad_Zuse/Neumann_vs_Zuse.html.
Zuse's work was in 1936, VN's in 1946. Not "a couple of decades"
difference, mind you. Also note that Zuse describes storing
instructions in memory, but this is not what Von Neumann promoted -- VN
promoted storing instructions in the same memory as data. Various
architectures (including the Harvard architecture) already existed
where code was stored in memory. It was just a different memory bank
than the data. It wasn't until Zuse's Z3, created in 1941, that the
concept of shared memory was employed by Zuse. Zuse and VN were quite
aware of one another at the time and, no doubt, bounced a lot of ideas
off one another. The bottom line, though, is that if you look at the
Z3 design, it has "almost all" of the features described by Von Neumann
in his ENIAC paper. Note, btw, that VN did not gain recognition for
creating a new archtecture, his paper simply described how the ENIAC
worked and then he went around promoting that design. A good
discussion of VN's "architecture" can be found on the Wikipedia at

http://en.wikipedia.org/wiki/Von_Neumann_architecture

As is usually the case, though, he who publishes first gets all the
glory, while he who sits on his invention is forgotten. John Von
Neumann made a big deal about giving talks and publishing papers on how
he thought computers ought to be designed. Hence, the generic
architecture was named after him and even though history is attempting
to correct itself with respect to who *invented* the architecture in
the first place, the name attached will always be the "Von Neumann
architecture". BTW, there was no plagiarism involved. Von Neumann
himself never claimed to have invented these ideas on his own, it's
just that he promoted it so heavily that he was given credit for it by
the press.

In any case, the shared memory concept was invented prior to Von
Neumann popularizing it. Though it wasn't that far in advance. If Zuse
had anything to do with such a design, he completely failed to let more
than a few people know about it. VN's big contribution was convincing
computer manufacturers to adopt the design.

BTW, various Harvard architectures *are* in common use today,
particularly in the microcontroller arena.
Cheers,
Randy Hyde

From: Alex McDonald on 3 Jan 2006 13:43
randyhyde(a)earthlink.net wrote:

> Hi Jonathon,
> Today, most people recognize the fact that Von Neumann got his ideas
> for a shared-memory computer from some peers that he was working with
> at the time (John William Mauchly and J. Presper Eckert?). There is an
> article about Zuse vs. VN here:
> http://irb.cs.tu-berlin.de/~zuse/Konrad_Zuse/Neumann_vs_Zuse.html.

Is the author Zuse's son?

Here's another wikipedia entry that seems relevant;
http://en.wikipedia.org/wiki/History_of_computing_hardware.

--
Regards
Alex McDonald
From: Ludwig Hähne on 3 Jan 2006 16:08
Alex McDonald wrote:
>> There is an article about Zuse vs. VN here:
>> http://irb.cs.tu-berlin.de/~zuse/Konrad_Zuse/Neumann_vs_Zuse.html.
>
> Is the author Zuse's son?

Yes. An interesting interview with him for all those who are able to
read German:

http://www.3sat.de/3sat.php?http://www.3sat.de/neues/dial/50434/

A problem Zuse had to face was that the Z3 was totally destroyed in WW2
and that it took decades to prove that he really built it. However, that
was more in competition with Howard Aiken and it is said that Zuse
honoured von Neumanns scientific merits (and probably had no problem
with this architecture being accredited to him).

Ludwig
 | 
Pages: 1
Prev: Check out POASM
Next: Blocks and compilers