Electronic Pioneers
Spring 2022
Introduction
A century after Babbage, in the 1930s and 1940s, there were multiple, independent efforts to build electronic computers. A strong case can be made that Konrad Zuse built the first modern computers, though wartime secrecy in Germany meant that researchers in the US didn't learn of his work until years later. The first paper below is an overview of the architecture of Zuse's Z1 and Z3 machines. The second paper is a report by John von Neumann and colleagues in 1946 as they were designing EDVAC, and summarizes the group's thoughts about the state of the art. Von Neumann was involved in the construction of several early machines in this country, and was one of the most influential researchers of the era. Read at least to the end of page 9 (page 16 of the PDF). The third paper is optional, but gives a little more historical context on Zuse in case you're curious.
Questions
As before, when reading these papers, don't get too hung up on the low-level details. I'm more interested in having you come away from the readings with the Big Picture. Here are some questions you might try to answer as you read:
-
What arithmetic operations did the Z3 support?
-
What kinds of numeric data could the Z3 process? Integers? Floating point? Both?
-
Did the Z3 use binary representations of values or binary coded decimal (BCD)?
-
In terms of its organization and capabilities, what are the most significant ways in which the Z3 differs from a more modern machine?
-
What is the "jewel in the crown of the Z3"?
-
What was the "main defect" of the Z3?
-
What arithmetic operations were supported by the machine von Neumann described?
-
How many data values was von Neumann planning to store? Why?
-
What are the "really decisive considerations" in "selecting a code"? What does that even mean?
-
What's an "unconditional transfer"? Do modern machines have them?
-
What's an "acoustic feed-back delay line"? What would it be used for?
-
What is von Neumann talking about when he refers to a "parallel machine"?
-
Does von Neumann propose using binary representations of values or binary coded decimal (BCD)?
-
What's von Neumann's beef with "so-called 'floating decimal point'" values?
-
Does the proposed machine have a built-in square root operation? Why or why not?
Papers
-
Raúl Rojas. 1997. "Konrad Zuse's Legacy: The Architecture of the Z1 and Z3". IEEE Ann. Hist. Comput. 19, 2 (April 1997), 5-16. DOI=http://dx.doi.org/10.1109/85.586067 [link]
-
Arthur W. Burks, Herman H. Goldstine, and John von Neumann. 1947. "Preliminary discussion of the logical design of an electronic computing instrument." (Second Edition) [link]
-
[Optional] Wolfgang K. Giloi. 1998. "Konrad Zuse: reflections on the 80th birthday of the German computing pioneer". SIGNUM Newsl. 33, 2 (April 1998), 11-16. http://dx.doi.org/10.1145/290590.290591