XB_0076 History of Science for CS
Index
Chapter 1 - Inventing the Computer .................................................................................................................1
Chapter 2 - The Computer Becomes a Scientific Supertool .............................................................................2
Chapter 6 - The Computer Becomes a Communications Platform — Incomplete ...........................................4
Chapter 13 - The Computer Becomes a Network .............................................................................................6
Chapter 14 - The Computer is Everywhere and Nowhere ................................................................................8
Chapter 15 - Epilogue : A Tesla in the Valley .................................................................................................10
Chapter 1 - Inventing the Computer
➤ENIAC
• Starting in 1943, the ENIAC had been designed and built in the University of Pennsylvania by Mauchly
and Eckert. It’s the first device to be electronic and programmable.
• The ENIAC was the only programmable electronic computer available for scientific use.
Programmability, its Design and Construction, and its Use
• ENIAC was not the first programmable computer but was the first to automate the job of deciding what
to do next after a sequence of operations finished. Early relay computers (e.g. Harvard’s Mark I) relied
on human operators to step in when decisions had to be made.
• The ENIAC was programmed by a team of women of the Moore School. They were employed to make
the ENIAC compute firing tables for the war.
• Data went in and out of ENIAC on punch cards since it was much faster than using paper tapes.
➤The EDVAC Approach
• By mid 1944, there already was a plan for the successor of ENIAC since it was so hard to operate;
switching between jobs was lengthy and error-prone.
• The EDVAC was centered around a new kind of memory, the delay line.
Von Neumann’s First Draft
• Mauchly and Eckert welcomed Von Neumann into their team for EDVAC. It would load programs into
main memory, which set the EDVAC apart from existing computer designs.
• The enormous influence of the EDVAC can be separated into three paradigms.
1. Hardware: All-electronic machine with large high-speed memory using binary number storage.
2. Von Neumann Architecture: Equivalence of brains and mechanical control mechanisms. Storage
and arithmetic was binary, using bits to encode information. Its structure was broken up into parts.
3. Modern Code: The flow of instructions and data mirrored the way humans performed scientific
calculations. Each instruction with an operation code was usually followed by parameters or an
address.
• Because the architecture really simple, more instructions were needed to do certain things. These
instructions took several steps to carry out. The basic cycle: fetch, decode, execute.
Early EDVAC-like Computers
• New computers has simple architectures, but making them took years. The biggest challenge was
producing stable memory.
• The Manchester Mark 1 was a stepping stone to a useful EDVAC-like computer in 1947. Instead of
delay lines, it used William tubes. A single tube stored much more, but was not always reliable since the
visible dots on the tubes could fade away.
• The EDSAC was the first operational full-scale computer in 1949 to implement all three of the EDVAC
paradigms. It became the basis of a scientific computing service.
Variation on a Theme
• Each computer project launched in the 1940s was an experiment, and designers tried out many variations
on the EDVAC theme. Successful ideas were widely copied, such as holding data in a matrix structure
and storing programs and data in addressable memory.
• EDVAC had grown into a four-address machine, with the fourth address specifying the location of the
next instruction to be read.
1
, • The EDSAC delivered one bit at a time, since it was serial. Most parallel-word computers used William
tubes, which could retrieve all the bits in a word of memory simultaneously.
• While Von Neumann was intrigued by Alan Turing’s concept of a “universal machine”, it didn’t shape
his design for the EDVAC.
Programming Tools
• General purpose computers can do many things. The disadvantage to that flexibility is that getting any
particular task done takes minutely detailed programming.
• New possibilities for automating program preparation founded the idea of assemblers. These assembled
subroutines and new code into a single executable program. It also translated each line into the
corresponding numerical instruction that the computer could execute.
• Symbolic and macro assemblers in the 1950s made assembly coding less tedious.
➤The Commercialization of Computing
The First Computer
• Eckert and Mauchly designed and built the Univac (1951) with their own company. It reflected Eckert’s
philosophy of loading the vacuum tube circuits conservatively and adding enough redundancy to ensure
reliable operation.
• The Univac was usable for business applications as well as scientific applications, because of the flexible
general outline of the EDVAC design.
• Univac started the era of large computers for what were later called “data processing” applications.
• The closest technology in widespread administrative use was punched card machines of IBM.
An Automated Utopia
• General Electric purchased the Univac for business administration purposes and it hoped to do more
than replace punched card machines. The age of atomization began with the Univac.
• The history of computing is the story repeated redefinitions of the nature of the computer itself, as it
constantly opens new markets, applications, and places in the social order.
Chapter 2 - The Computer Becomes A Scientific Supertool
➤The First Scientific Computers
The 701 : Defense Calculator
• In May 1952, IBM announced the 701, also known as the Defense Calculator. It was comparable to the
Univac but memory reads were faster because it retrieved all the bits at once.
• The 701 was used for weapons design and logistics in the business world. It also predicted an election.
The IBM 704 and its Successors
• By 1956 IBM installed more large computers than Univac. Much was owed to the 704, which was
announced in 1954 as the improved successor of the 701. It had three key improvements:
- Core memory instead of William tubes
> Each core stores one bit of information, by magnetizing it one of two possible ways.
> Small and nonvolatile, retains information when power is removed.
> Provides random access, which made it stand out against delay lines and drums.
- Hardware support for floating-point arithmetic, simplifying planning and programming.
> Allowed faster and simpler programs but makes the processor more complicated and expensive.
- Three index registers to simplify programming
• The Whirlwind memory used a three-dimensional array for core memory, improving the original.
Interrupts and Channels
• Magnetic tapes and drums supplied information much faster than punched cards, but not nearly quickly
enough to keep processors busy.
• By the mid 1950s, computer manufacturers introduced interrupts. When printers were ready they could
ask for an interrupt and jump to a subroutine to handle the situation.
• SPOOLing was an application of interrupts in the IBM 7070, which let programs write their printed
output to a tape file, so the computer could move on to its next job.
• The 709, introduced in 1958, added an innovation that has defined the mainframe of the computer ever
since: input and output channels for high data throughput.
• Without interrupting the main processor, it could print a file, copy data from one tape to another, or load
a file into memory.
2
