The ENIAC - Wittenberg University

advertisement
The Architecture of the ENIAC
The Electronic Numerical Integrator and Computer
Brian J. Shelburne
Dept. of Math and Comp Sci.
Computer architecture deals with the functional
behavior of a computer system as viewed by a
programmer.
Overview
Background and Origin of Project PX
The Decade Counter
Representing Numbers on the ENIAC
Units of the ENIAC
The Accumulators
A Sample Calculation
Sequencing Accumulators
The Master Programmer
Epilogue
Background and Origins of Project PX
The ENIAC (Moore
School of Electrical
Engineering at
UPenn) was funded
by the US Army
during WWII to
compute ballistic
tables.
Vannevar Bush’s Differential Analyzer
The Moore School
had one of the
existing three
Differential
Analyzers , the
other two being
at MIT and the
Army’s BRL at
Aberdeen MD
http://www.library.upenn.edu/exhibits/rbm/mauchly/img/5women.jpg
The People
John W. Mauchly
1907 – 1980
J. Presper Eckert Jr.
1919 - 1995
The People
H.H. Goldstine
1913 - 2004
Project PX – the “ENIAC” - proposed 1942,
contract signed 1943, delivered 1946
http://coolphotobench.com/images/Cars/adele-goldstine.jpg
"ENIAC OFFICIALS" From left to right: J. Presper Eckert, Jr., Chief Engineer; Professor J. G. Brainerd, Supervisor; Sam
Feltman, Chief Engineer for Ballistics, Ordnance Department; Captain H. H. Goldstine, Liaison Officer; Dr. J. W.
Mauchly, Consulting Engineer; Dean Harold Pender, Moore School of Electrical Engineering, University of
Pennsylvania; General G. M. Barnes, Chief of the Ordnance Research and Development Service; Colonel Paul N.
Gillon, Chief, Research Branch of the Army Ordnance Research and Development Service.
http://explorepahistory.com/kora/files/1/2/1-2-1573-25-ExplorePAHistory-a0l4o2-a_349.jpg
Project PX: The ENIAC Project: 1943-1946
30 units (40 panels) including
20 Accumulators capable of storing a signed 10
digit decimal number (200 digits of storage)
Capable of 5000 additions/second (200 msec)
~18,000 vacuum tubes
30 tons - 1800 ft2 - 150 KW power
Cost ~$500,000.00
Functional
Diagram of
the ENIAC

  















• Accumulator
× Multiplier
÷ Divider/Square Rooter
c Constant Transmitter
M Master Programmer





c
c
M
M
c
The Decade Counter:
the basic computational unit
Representing Numbers
A digit was stored in a 10 stage ring counter circuit
(10 two-tube flip-flops) – a.k.a. “decade counter”.
only one of the ten flip-flops being in an
abnormal state (thus indicating the digit value)
If the decade counter received a “pulse” the
position of the abnormal state advanced to the
next position mod 10 (with a carry out pulse!)
Ten decade counters plus an 11th two-state counter
for the sign allowed a signed 10 digit decimal
number to be stored per accumulator.
The Decade Counter: 7 + 4 = 1 plus carry
…
1
0
9
1
0
9
2
8
2
8
3
7
3
7
4
5
6
4
4
5
6
Representing Numbers on the ENIAC
– Doing Arithmetic
Ten Digit 10’s Complement Notation
P 9 999 999 999
<- positive integer
a sign plus ten digits
M 9 999 999 999
<- negative integer
In 10’s Comp. Not. the sign digit is given a value
P = 0 , M = -1010
So -1 is represented by M 9 999 999 999,
-2 by M 9 999 999 998 etc.
Simple Example: Three Digit 10’s Complement Notation
Sign “digit” has value: P = 0; M = -1 but it’s
“position” has weight is 1000
Therefore - 124 is represented by M 876
Ck: -1000 + 876 = -124
Addition “works” with 10’s Comp. Notation
257 → P 257 → 0257
- 124 → M 876 → 1876
133
P 133
0133
throw out carry  outs
It’s easy to negate numbers – form the 9’s
complement of each digit and add 1
Example: -257 → 742 + 1 → 743
Check: Add 743 to 257 and throw out the carry
0 257
+1 743
0 000
If you negate M 743 → P 256 + 1 → P 257
Note: Electronically it was easy for the ENIAC to
negate a number
Unit of the ENIAC
30 Units – 40 Panels
20 Accumulators
Multiplier (3 panels)
Divider/Square Rooter
3 Function Tables (2 panels each)
Constant Transmitter (3 panels)<- Card Reader
Card Punch Unit (3 panels) -> Card Punch
Initiating Unit/Cycling Unit (2 panels)
Master Programmer (2 panels)
Functional
Diagram of
the ENIAC

  















• Accumulator
× Multiplier
÷ Divider/Square Rooter
c Constant Transmitter
M Master Programmer





c
c
M
M
c
The
Accumulators
The Accumulator …
… could store a signed 10 decimal digit number.
When it received a control pulse it could
1. Receive a 10 digit signed number or
2. Transmit additively or subtractively its contents
over data lines
3. Repeat the above a fixed number of times
depending on the setting of a console switch
4. Send a control pulse (CCP) when it was done
Accumulators could be daisy chained together to
perform sequence of calculations
View of ENIAC accumulators. Standing: Ester Gerston;
crouching: Gloria Ruth Gorden. (US Army Photo; http://ftp.arl.army.
mil/ftp/historic-computers/png/eniac4.png.)
• http://www.library.upenn.edu/exhibits/rbm/mauchly/img/eniac4.jpg
A Sample Calculation
Example taken from The Electronic Numerical Integrator and Computer (ENIAC) by H.H. Goldstine
and Adele Goldstine; MTAC 1946; Reprinted in IEEE Annals of the History of Computing, Vol 18,
No. 1 1996, pp. 10 - 16
An Sample Calculation
Program a sequence of operations that calculate
values of n, n2 and n3 for integers between 1 and
1400 printing every 7th set.
Note:
(n+1)2 = n2 + 2n + 1
(n + 1)3 = n3 + 3n2 + 3n + 1
Example taken from The Electronic Numerical Integrator and Computer (ENIAC) by H.H. Goldstine
and Adele Goldstine; MTAC 1946; Reprinted in IEEE Annals of the History of Computing, Vol
18, No. 1 1996, pp. 10 - 16
Sample Calculation: Assume AC18 = n; AC19 = n2; AC20 = n3
n
n2
n3
1
Sample Calculation: A1: AC19 transmits 3X; AC20 receives 3X
AC18 = n; AC19 = n2; AC20 = n3 + 3n2
n
A1
n
2
n  3n
3
2
A2: AC18 transmits 3X; AC19 & AC20 receive 3X
AC18 = n; AC19 = n2 + 2n; AC20 = n3+3n2+3n
n
A2
n 2  2n
n3  3n2  3n
A3: CT transmits 1; AC18, AC19,AC20 receive 1X
AC18 = n + 1; AC19 = n2+2n+1; AC20 = n3+3n2+3n+1
n 1
A3
n  2n  1 n  3n  3n  1
2
3
2
The Master Programmer
Four panels of ENIAC, with Betty Jennings and Frances Bilas (right) arranging the
program settings on the master programmer, 1946
http://www.library.upenn.edu/exhibits/rbm/mauchly/img/7pansen.jpg
The Master Programmer consisted of 10 units
each consisting of ten decade counters and five
6-stage stepper-counters
It coordinated the 20 accumulators (and other
units) providing sequencing and loop control.
It could emit a control pulse to initiate a
sequence of accumulator operations counting
the number of times that sequence occurred.
When the count was reached it could initiate the
next sequence of accumulator operations
There was a work-around for conditional
branching
Sample Calculation: The Initial Step – A1
A1
Sample Calculation: CT -> A5
Sample Calculation: Decision – Print or Continue? A1
A1
Sample Calculation: A4 - At 200 end else continue
Sample Calculation
Epilog : 1946 – 1955
and beyond
The ENIAC was completed too late to produce ballistic tables
for the Army in WWII
It’s first real test was a secret and complex calculation for a Los
Alamos labs group working on the hydrogen bomb (late
1945? – rumored to require 1,000,000 punch cards)
The ENIAC was announced to the public on 2/14/46
On 11/9/46 it was shut down and transferred to the Army’s
BRL at Aberdeen MD. It was restarted 7/29/47
While at the BRL the ENIAC was redesigned so that it could be
programmed via its function tables; its initial 60 instruction
set (A Logical Coding System Applied to the ENIAC - 1947)
was later increased to 100 instructions.
In 1949 it computed π to 2000+ digits a record that held for 5
years!
It was finally turned off on 10/2/55
Units of the ENIAC can be seen at UMich, UPenn, and the
Smithsonian
John von Neumann
1903 - 1957
How von Neumann became interested in the ENIAC
and in computers in general is interesting. In August
1944 while waiting at a R.R. station H.H. Goldstine
introduced himself to von Neumann. He writes
“I was waiting for a train to Pennsylvania on the
railroad platform at Aberdeen when along came
von Neumann. Prior to that time I had never met
this great mathematician but I knew much about
him of course and had heard him lecture on several
occasions. It was therefore with considerable
temerity that I approached this world famous
figure, introduced myself, and started talking.”
“Fortunately for me von Neumann was a warm,
friendly person who did his best to make people
feel relaxed in his presence. The conversation
soon turned to my work. When it became clear
to von Neumann that I was concerned with the
development of an electronic computer capable
of 333 multiplications per second, the whole
atmosphere of our conversation changed from
one of relaxed good humor to one more like the
oral examination of the doctor’s degree in
mathematics”
Von Neumann had no direct impact on the design of the
ENIAC
In 1945 he wrote “First Draft of a Report in the EDVAC” – a
paper on stored program computers. Today it is understood
that the ideas and proposals for the EDVAC (Electronic
Discrete Variable Automatic Computer)represent the
collective thinking of the ENIAC team.
During the summer of 1946 the “Moore School Lectures
“(7/8/46 – 8/30/46) disseminated information on the
EDVAC design. As a result ….
Cambridge University’s EDSAC (Electronic Delay Storage
Automatic Computer) computer went operational 1949
The EDVAC was finally completed in 1952
In 1946 along with A. Burks and H.H. Goldstine , vonNeumann
authored “Preliminary discussion of the logical design of an
electronic computing instrument”
Von Neumann and IAS Machine - 1951
Unlike the
EDVAC type
machines the
IAS machine
was parallel
using Williams
tube memory
instead of the
EDVAC acoustic
delay line
memory
Eckert & Mauchy
John Mauchy and J. Presper Eckhart foresaw the
commercial potential for the electronic computer
In 1946 both left the Moore School and formed the
Eckert-Mauchly Computer Corporation to build
an EDVAC type computer (UNIVAC-I)
EMCC was bought out by Remington Rand in 1950,
which merged with Sperry Corp -> Sperry Rand
Corp. which today is Unisys.
The first UNIVAC was delivered to the US Census
Bureau in 1951
In 1973 the ENIAC patent was voided (Honeywell
Inc vs. Sperry-Rand Corp.)
Women of the ENIAC
Adele Katz (Goldstine) -> authored Report on the ENIAC (1946)
Ruth Rauschenberger (Ammlung)
Lila Todd (Butler)
Kathleen McNulty Mauchly Antonelli
Frances Elizabeth (Betty) Snyder (Holberton)
Betty Jean Jennings (Bartik)
Marlyn Wescoff (Meltzer)
Frances Bilas (Spence)
Homé McAllister (Reitwiesner)
Marie Bierstein (Malone)
Willa Wyatt Sigmund
Ruth Lichtermann (Teitelbaum)
Source: “The Women of ENIAC” by W. Barkley Fritz, IEEE Annals of the History of
Computing, Vol. 18, No. 3 1996
ENIAC at BRL: (left to right) Homé McAllister, Winifred (Wink) Smith, George
Reitwiesner, and Ruth Lichterman
Sources & References
http://www.library.upenn.edu/exhibits/rbm/mauchly/jwmintro.html
The Computer from Pascal to von Neumann by H.H. Goldstine; Princeton
University Press, 1972.
From ENIAC to UNIVAC: An Appraisal of the Eckert-Mauchly Computers by
Nancy Stern; Digital Press, 1981
“The ENIAC: History, Operation and Reconstruction in VLSI” by Jan Van der
Spiegel, James F. Tau, Titiimaea F. Ala’ilima, Lin Ping Ang printed in The
First Computers History and Architectures edited by Paul Rojas and Ulf
Hashagen; MIT Press, 2000
Google ENIAC \ Videos
Download