Professional Documents
Culture Documents
The module in the background is exactly the same as one in the foreground. but it has only been used on Earth.
The Museum's prototype computer ran at Draper Labs and was used to test the routines for the in-flight machines.
In space all of the components had to be totally "potted" to insure that all the parts would stay firmly in place
and remain uncontaminated during space flight.
The Apollo Guidance Computer,
shown on the left, was responsible
for the guidance, navigation. and
control computations in the Apollo
space program. The AGC was the
first computer to use a n integrated
circuit logic and occupied less than
1 cubic foot of the spacecraft. It
stored data in 15 bit words plus a
parity bit and had a memory cycle
time of 11.7 microseconds. utilizing
2,000 words of erasable core mem-
ory and 36,000 words of read-only
memory. The frame is made of
magnesium for lightness and de-
signed to hermetically seal the
components.
The interface with the astronauts
was the DSKY, shown on the riaht.
It used digital displays and coh-
municated with the astronauts
using the verb and noun buttons
visible in the photograph and two
digit operation and operand codes.
The AGC rope memory A set of status and caution lights
is shown in the top left corner
of the DSKY.
The read-only memory of the com- The AGC and DSKY are on display
puter consisted of six rope memory in the Four Generation Gallery.
modules, each containing 6,000
words of memory. This special type
of core memory depended on the
patterns set at the time of manufac-
ture. Its sensing wires were woven
into a set pattern information. It
had five times the density and was
far more reliable than the coincident
current core memory used for eras-
able storage in the computer. Being
unalterable. it also provided a
greater incentive for error-free
software development.
The AGC rope memory is on dis-
play in the primary memory case.
The Apollo 11lunar landing had wrong mode during the lunar land- capacity. However, no commercial
an anomaly which attracted public ing phase, wasting computer mem- computer could claim to match the
attention. The computer in the LEM ory cycles. The computer software power consumption and space
signalled a restart alarm condition was responding to overloads as characteristics of the AGC.
several times during a very critical designed.
Excerpted from a n Illustrated
period prior to touchdown. This fact This incident triggered a news Lecture, June 10, 1982, by Ben Gold-
was broadcast to the public and berg. The video-tape is archived by
those who knew its significance brief in Datamation in October, 1969,
faulting the computer design for be- The Computer Museum.
were close to a state of panic. After
ing too slow. It rightfully claimed
analysis, it was determined that the that there were a numbe; of mini-
alarms were a n indication to the computers, including the PDP-11,
astronauts that the computer was that were at least a n order of mag-
overloaded and was eliminating low nitude faster. In the eight years since
priority tasks from the waitlist.
The overload resulted from the
rendezvous radar being set in the
1 the initiation of the A ~ o l l oDroaram
commercial technology h a i fay sur-
passed that of the Apollo design and
1
The Apollo Guidance Computer .
David Scott Astronaut for the Gemini 8, Apollo 9, and Apollo 15 missions.
In 1963 when NASA was con- return to Earth. Between 1963 and I called up MIT and asked if they
ducting the selection of the third 1969, with the flight of Apollo 9 this could program the computer to give
group of astronauts for the U.S. was accomplished. I stayed in the the vehicle a satisfactory orbit rate.
space program, I had just received spacecraft while Rusty Schweickart They answered, "Of course, which
a graduate degree at MIT and fin- and Jim McDivitt got in the lunar way do you want to go and how
ished test pilots school. My inter- module and went out about 60 fast?". In a matter of a couple of
ests and the program's need for a miles away. The computer behaved days we had a program and a
user to interact with the design of flawlessly during our first success- simulator that automatically drove
the guidance computer at the MIT ful rendezvous in space. a spacecraft at perfect orbit rate.
Instrumentation Lab was a good fit. We got into flight with very little
I was part of those discussions Another assignment for Apollo chance to practice or verify, but
whether to use analog or digital 9 was to take the first infra-red pho- we put on the cameras and the
controls that Eldon described. tographs of the Earth from space. results were perfect.
To do this, a large rack of four cam-
The MIT Interface eras was mounted on the space-
When I was studying at MIT, craft. Since they were fixed to the Potential Computer Failure
the ability to rendezvous in space spacecraft, the vehicle itself had During the development pro-
was a n issue for debate. It wasn't to track a perfect orbit such that cess we ran many simulations of
clear whether it was possible to de- the cameras were precisely vertical in-flight computer operations with
velop the mathematics and speed with respect to the surface that they particular concern for in-flight fail-
of computation necessary to bring were photographing. During simu- ure. But in the 10 years that I spent
two vehicles together at a precise lations it was determined that in the program there was never a
point in space and time-a critical manual orbit procedures would real computer failure. Yet, people
issue for the Apollo missions suc- be inaccurate. We were at a loss. often wonder what a computer fail-
cessful landing on the moon and About two weeks before the flight ure would have meant on a mis-
sion. It would have depended on
the situation and the manner in
which the computer failed. We
probably would not have expired,
but there were some parts of the
mission in which a computer fail-
ure would have been especially
compromising. Navigation was not
necessarily time critical but the lu-
nar landing was very time critical.
You could have a situation during
a Iunar landing in which, if the
computer failed, the engine would
be driven into the ground. Unless
the astronaut could react quickly
enough to stop it, the Lunar Module
could have been flung on its side.
Chances are that the astronaut
could prevent such a n event by
switching to manual control of the
vehicIe. It must be remembered
that the computer had been de-
signed to be a s reliable as pos-
sible and the astronauts had a
great amount of confidence in
the machine.
8-
The Apollo 9 Crew, (from left) Dave Scott,
I Command Module Pilot; Jim McDivitt, Com-
mander; Rusty Schweickart, Lunar Module
Pilot. Apollo 9 w a s launched on March 3,
1969. The first separation and rendezvous
of the Lunar and Command Modules w a s
carried out in Earth Orbit on this flight.
8 GALLERY TALK
Whirlwind Before Core imal and octal. My boss, Charlie refer to a n address above where
Adams, was concerning himself you were, as opposed to everything
Reminiscences of Jack Gilmore with that cmd so it became my job going below. The two pass assem-
to write the assembly program. bler came out of all that. I have a
In October, 1950, I joined the I'm fairly certain that if it is not recollection of Charlie Adams and
Whirlwind team. At that time the the first, it is one of the very first I briefing IBM's Nat Rochester
first thirty-two registers of toggle assembly programs ever written. on how to produce symbolic
switch memory were working. The only one that I know of that addresses.
The four variable flip-flop registers predates it was Wilkes' 'Load
could be assigned to any one of the and Go' on the EDSAC. The Ph.D. candidates who
thirty-two addresses. They were needed to use the Whirlwind
In September, 1951, John Carr, really didn't know how to run the
able to demonstrate small mathe-
later Chairman of Duke's Computer machine. There were full scale
matical programs such a s the
Science Department, and I wrote a electronic technicians who knew
bouncing ball problem or solve
document that explained how peo- how to bring it up, and most of the
simple differential equations. The
ple could actually use subroutines systems programmers like myself
first memory consisted of electro-
in conjunction with assembly pro- knew how to do it, a s well a s some
static storage tubes totaling 256
grams, so that they didn't have to of the engineers. It was a fairly
locations. We felt really rich with
write all the various utilities. Peo- routine procedure so I went to
a full 256 variable registers to write
ple could write their programs in a Charlie Adams and suggested that
our programs. We calculated the
relative fashion and then we would I could train two people right out
operation in the octal address and
give them the library of subroutines of high school to be computer
then looked up what was then
and they'd actually pick out the operators if I had enough funds to
called the sexidecimal conversion
tapes that they needed. We'd then hire them for one year. Jay Forrester
number (later the term hexadeci-
string the tapes together and provided the funds and I went out
mal was used). We had a little load
literally make a copy not only of to two local high schools and asked
program in the 32 registers and that
their program but also of the sub- for students that were college mate-
bootstrapped the programs up into
routines. All of those would be rial but didn't have the money for
the memory in order to run them.
pulled in through the bootstrap college. I hired Joe Thompson from
The first thing that we were program and it would run. This Boston Technical High School
very anxious to do was to get a n was the indirect birth of the sym- (shown sitting down in the photo-
assembly program that would bolic address. The thing that we graph) and Bill Kyle from Boston
allow us to be able to write our pro- discovered, I think I actually dis- English. Within four or five months
grams using mnemonic symbols covered it, was that when we ran they were competent operators,
and expressing the numbers in dec- the tape through twice, you could and Joe stayed on to complete his
At the Opening and first board meeting. many of the Board members had their first tour of the Museum. The old
hands. such a s Gordon Bell. each had their assignment to explain and guide a new board member through the
exhibitions. Gordon is giving Senator Paul Tsongas and his aide Andy Bagley a n overview using the self-guiding
brochure now available.
riRCHIVES AND LIBRARY INTERN PROGRAM Apollo Guidance Computer in
this issue a n d will wqrk on a com-
This fall, Gregor Trinkaus- This summer, eight interns prehensive exhibit of the Polaris
Randall will join the staff as archi- worked a t the Museum on various a n d Apollo Guidance Computers.
vist and librarian. Gregor will work special projects. Students a r e in-
to establish a n archives a n d library vited to apply for paid internships David Bromfield, a Senior
for scholarly use. a t the Museum for any period u p to at MIT in Business Management,
six months. The number of students is the Acting Business Manager.
An Archives Advisory Commit- As such, he h a s set up the initial
will vary according to the number
tee met for the first time on June 9 accounting systems for the
of activities going on at any time.
and will meet again in May of 1983. Museum.
Internships will complement the
Committee members will assist
various operational sections of the Roberto Canepa, a junior in
Gregor in the development of a
Museum. The listing of this sum- electrical engineering a t Carnegie
comprehensive computing history
mer's interns a n d their activities Mellon University, h a s been re-
archives. Those interested in par-
indicates the wide range of pos- building the number sieves built by
icipating should contact Gregor
sibilities. Professor Derek Lehmer of Berkeley.
kinkaus-Randall a t the Museum.
His two biggest challenges are t h e
The Library a n d Archives Beth Parkhurst started as a n recreation of a bicycle chain ma-
will be located on the lobby floor intern in the summer of 1981. She chine from photographs taken
of the Museum a n d should be open is now a Research Assistant a t the in the 1920s a n d the restoration of
in the spring of 1983. The collection Museum, spending one day a week the gear machine exhibited a t the
will include documents relevant during the school year while s h e 1932 Chicago World's Fair.
to exhibited artifacts. With only works to complete her Ph.D. in the
a small budget for acquisitions, Allison Stelling, a junior a t
American Civilization program at
the Library a n d Archives will be Harvard majoring in History a n d
Brown University. Computer Science, h a s concen-
dependent upon donations of ,
dehccessioned materials from
company a n d private libraries.
Andy Kristoffy, a junior in Com-
puter Science a t the University of I trated on doing research to gather
photographs for our collection.
The
Computer
Museum
One Iron W q
Marlboro
Massachusetts
01752