Go to Previous,
Go to Next
D. STORAGE
An extremely diverse and dynamic field of interest in the
study of computing systems is the subject of storage devices.
many ingenious devices, utilizing the ability of various material
media to store energy, have been devised during the past several
years. Early forms of storage involved mechanical deformation
of material media. These are exemplified by cams, springs, gears,
music box cylinders, perforated player piano rolls, code wheels
and perforated paper tape. All these storage devices required
the movement of large masses of material and consequently long
access time was inherent. The capacity in terms of stored in-for-
mation per unit volume of material was very low.
During World War II, the search for more rapid access storage
devices led to the use of the vacuum tube as a storage device. The
two states of conduction and cut-off permit information storage.
This system, as is used on the ENIAC, proved effective from an
access time consideration, however the system `was extremely bulky
and required thousands of tubes for a 20 word storage unit.
Chronologically the next development was the acoustic delay
lines of mercury and quartz. A transducer at each end of a length
of these materials permits enery conversions and allows the storage
of information in the form of high frequency (8 megacycles/sec)
pulse packets. The information is continuously recirculated.
Information is inserted or read out through the use of standard
gating techniques. Among the computers utilizing an acoustic
mercury delay line one finds are the DYSEAC, EDVAC, ELECOM 125,
FLAC, MIDAC, RAYDAC, SEAC and UNIVAC. The TECHNITROL-180 computer
utilizes an acoustic quartz delay line. Other types of delay
lines used for storage of information are the nmgnetostrictive, as
used in PEGASUS and the electromagnetic, as used in the FERRANTI
computers. Although in operating principle there is no difference,
it is necessary to make a distinction between a delay line used
in a storage loop in which information is continuously circulated,
and a delay line used only for purposes of timing the arrival
of information at selected points for performing various logical
operations. In the latter, the function is delay, or very
temporary storage, rather than full or permament storage. Since
delay lines store information serially as a train of electrical
or sonic pulses, average random access time `was limited to
somewhat more than half of the time length of the delay line.
The search for faster access yielded the use of the electro-
static or cathode ray tube as a storage device. The material
media in motion was now limited to electrons in beams and on
charged areas on the screen of a cathode ray tube. These charged
areas behaved somewhat like an array of charged capacitors. Selection
of storage locations was efficiently performed by an easily de-
flected pencil or beam of electrons which `was used for storage
and interrogation.
The electrostatic storage system, with its inherent problems
of high accelerating voltage sources, screen imperfections and
other tube failure problems, is gradually yielding to the magnetic
core. A 52 x 52 array of cores, which constitutes a storage plane,
may measure only a few inches on each side. The cores are placed
at the intersection of the wires of a mesh, and a third winding
is threaded through all cores for sensing. The storage takes
place in the form of magnetically oriented molecular or atomic
dipoles which retain their orientation upon removal of the magnetiz-
ing force. Many manufacturers intend to provide computing systems
with large capacity core storage units, however, not many large
capacity core units are actually operating. Progress in the field
of core storage and switching has not been as rapid as `was originally
anticipated. The use of the tried and tested delay lines and the
cathode ray tube will probably be quite prevalent for several years.
Table V shows the approximate relative order of high speed storage
access times for the various media used In computing systems.
It must be emphasized that the question of precisely what constitutes
access time cannot easily be resolved unless a common understanding
as to the definition is reached. In the usual sense one may consider
access time as the elapsed time which transpires between the
initiation of a command to transfer an item from one address in
the storage to another designated register and the complete arrival
of the item. In many systems, particularly serial storage units,
access time depends upon the tlJne location of the word in the
serially circulating group of words at the instant the transfer
command is initiated. For this and other reasons, much
misunderstanding can arise In the consideration of access time.
The data presented in Table V should therefore be considered to
be approximate.
The capacity of high speed storage units has risen during
the past few years as rapidly as access time has diminished.
Table VI shows the capacity of high speed storage units in terms
of number of words and word length. In arriving at the approximate
relative order of capacity shown in the table, some problems arose
when considering the storage of signs. In addition, it was felt
that since a sonic or electric delay line, and electrostatic
storage unit, or a magnetic core unit ail stored information in
a binary fashion from the point of view of the medium, k bits/dec digit
was used to determine the storage capacity of a decimal machine,
even though the storage medium actually is not used as efficiently
as in a straight binary machine.
Rapid access storage of a limited capacity must be supported by
a large capacity for a well balanced storage system. This permits
the transfer of large blocks of information from the rapid access
storage unit to the large capacity storage unit for use at another
location or time in the computation process. The storage device
must be far more rapid than punched cards or paper tape. The most
prevalent device for auxiliary storage is the magnetic drum. The
access time for large blocks of information is of the order of
tens of milliseconds. Many computing systems, particularly those
in which reiteration is not too high, utilize magnetic drums as
the primary storage unit. Examples of these types of machines are
the IBM 650, ADEC, ALWAC III, CALDIC, CIRCLE, and NCR-CRC-102A-D.
Several systems utilize large capacity drum units for commercial
applications such as payroll, stock inventory, and personnel records
where access times of the order of microseconds are not required.
Examples of these systems are the FERRAHTI MARK I, TIM II, MAGNEFIIE D,
MAGNETRONIC RESERVISOR, and the LOGISTICS COMPUTER.
Table VII shows the capacity and access times of various
drum storage systems currently in use.
TABLE V
ACCESS TIME OF HIGH SPEED STORAGE UNITS
ACCESS TIME STORAGE COMPUTING
MICROSECONDS MEDIUM SYSTEM
------------ ------ --------
8 CRT MANIAC-II
8-16 CRT MANIAC
8 CRT NORC
8 MC UNI-SCI (ERA-ll03A)
8 MC WHIRLWIND-I
10 MC FER MARK-II
10 CRT NAREC
12 CRT IBM- 701
12 MC IBM- 704
12 CRT SEAC
12 CRT UNI-SCI (ERA-llO3)
15 MC JOHNNIAC
16 CRT SWAC
17 MC IBM-705
18 CRT ILLIAC
18 CRT ORDVAC
20 CET ORACLE
20 MC RCA BIZMAC
25 CRT IAS
40 MC UNIVAC-II
48-384 DL DYSEAC
48-384 DL EDVAC
48-384 DL FLAC
48-584 DL TECH-180
96 MC IBM-650
192 DL MIDAC
200 VT ENIAC
216 DL SEAC
220 MC IBM-608
505 DL RAYDAC
400 DL UNIVAC
500 VT IBM-604
520 VT IBM-607
600 MC ENIAC
760 VT IBM-CPC
1,500 VT NCR-503
7,000 MC TIM-II
Key to Symbols CRT Electrostatic Storage-Cathode Ray Tube
MC Static Magnetic Storage-Magnetic Core
DL Sonic or Electric Delay Line
VT Vacuum Tube
TABLE VI
CAPACITY OF HIGH SPEED STORAGE UNITS
CAPACITY SYSTEM CAPACITY SYSTEM
-------- ------ -------- ------
Words Digits/word Words Digits/word
97,500-12 dec LARC 1,024-36 bin UNI-SCI (ERA 1105)
12,288-48 bin MANIAC-Il 2,048-16 bin WHIRLWIND-I
10,000-11 dec+sign UNIVAC-Il 4,096-alpha-num RCA BIZMAC
12,288-56 bin UNI-SCI (ERA 1103A) 512-48 bin TECHNITR0L-l80
8,192-56 bin IBM-704 512-45 bin DYSEAC
4,096-40 bin JOHNNIAC 512-45 bin FLAC
2,000-16 dec NORC 512-45 bin MIDAC
20,000-alpha-num IBM-705 512-45 bin SEAC*
2,048-56 bin IBM-701 768-20 bin FERRANTI MARK-I
10,000-aipha-num IBM-702 256-57 bin SWAC
1,000-12 dec UNIVAC 55-59 bin PEGASUS
1,024-45 bin NAREC 40- 9 dec IBM-608
1,024-44 bin EDVAC 120-10 dec+sign ENIAC
1,152-56 bin RAYDAC 100-10 dec+sign ELECOM-125FP
4,096-10 bin FERRANTI MARK-II 60-10 dec+sign IBM-650
1,024-40 bin lAS 57- 5 or 5 dec IBM-607
1,024-40 bin ILLIAC 10-12 dec-var TIM-II
1,024-40 bin MANIAC 8- 9 dec NCR-303
1,024-40 bin ORACLE 9- 5 or 5 dec IBM-CPC
1,024-40 bin ORDVAC 9- 5 or 5 dec IBM-604
*SEAC utilizes 512 word-45 bit CRT Storage and 512 word-45 bit DL Storage.
Above Table is in approximate order of binary equivalent
capacity.
TABLE VII
MAGNETIC DRUM STORAGE SYSTEMS FOR DIGITAL COMPUTERS
CAPACITY SYSTEM
Words Digits/word-------
----- -----------
1,000,000 - 12 dec LARC
52,768 - 40 bin FER MARK-I
16,584 - 64 bin MELLON INST-DIG
26,000* - 10 dec TIM-II
8,000 - 21 dec MAGNEFILE-D
60,000 - 10 bin FER MARK-II
16,584 - 56 bin IBM-704
16,584 - 56 bin UNI-SCI (KRA-1103A)
56,848 - 16 bin WHIRLWIND-I
12,000 - 40 bin JOHNNIAC
20,000 - 6 dec MODAC-404
10,000 - 44 bin OARAC
10,000 - 10 dec CALDIC
10,000 - 10 dec ELECOM-120A
10,000 - 10 dec ELECOM-125
10,000 - 10 dec WHITESAC
10,000 - 40 bin MANIAC
16,584 - 24 bin UNI-SCI (ERA-llOl)
60,000 - alphan IBM-702
10,000 - 8 dec ORDFIAC
8,192 - 56 bin IBM-7Ol
6,144 - 44 bin MIDAC
4,000 - 16 dec ADEC
5,000 - 12 dec OLIVETTI-GBM
5,500 - 10 dec UDEC-Il
4,608 - ~ bin EDVAC
10,000 - 5 dec MDP-MSI-5014
5,000 - 10 dec MODAC-410
8,192 - 24 bin UNI-SCI (ERA-ll02)
5,500 - 9 dec UDEC-I
4,096 - 46 bin CIRCLE
4,608 - 40 bin PEGASUS
15,000 - 5 dec MAGETRONIC RES
4,000 - 10 dec DATATRON
4,000 - 10 dec READIX
5,840 - 10 dec MINIAC
4,000 - 58 bin ADEC
4,096 - 37 bin SWAC
4,096 - 54 bin ALWAC-III
4,096 - 52 bin LGP-30
4,040 - 8 dec MAGNAFILE-B
2,500 - 10 dec PENIISTAC
2,000 - 10 dec IBM-650
2,048 - 40 bin IAS
2,048 - 40 bin ILLIAC
1,556 - 45 bin NAREC
2,048 - 52 bin ALWAC-III
2,160 - 29 bin BENDIX-G15
1,024 - 50 bin WISC
1,024 - 42 bin NCR-CRC-102D
1,024 - 41 bin NCR-CRC-102A
1,024 - 9 dec NCR-303
1,984 - 17 bin HUGHES AAC MOD-III
4,092 - 8 bin RCA BIZMAC
1,200 - 25 bin HAL RAY BROWN
1,000 - 6 dec BAEQS
650 - 8 dec BENDIX-D12
200 - 20 dec MONOBOT-VI-MU
512 - 50 bin ELECOM-l00
100 - 20 dec MONROBOT-IlI
100 - 12 dec BUR-El0l
100 - 10 dec ELECOM-50
16 - 20 dec BAR-CAL DEC DIG
*Variable up to 76,000 words.
The characteristics of a storage device, namely, capacity and
access time are two aspects of a storage system which come under
consideration when designing or using a machine. The designer
can utilize either characteristic in accordance with his design
decisions. The user of a system, at times, can trade capacity for
access in the sense that he can accomplish an equivalent amount
of computation with a large capacity-long access time system as
with a small capacity - short access time system. There are
limits, in the sense that when access approaches the order of
milliseconds, computation is seriously slowed down. Since large
capacity and short access time are features to be desired, let us
examine a quantity determined by the expression:
Log10 (Capacity in binary digits/access time in seconds)
Historically, for old storage devices as music boxes and signal
coding devices, this ni.miber is of the order of two to three.
Relay
storage units have a number of the order of four. Tube registers
of the HNIAC tube accumulator storage units raised this figure to
6.ll. Magnetic Drum storage units operate in the region of 5.5
to 6.5. Acoustic delay line storage systems brought this figure
to 5.7. The cathode ray tube storage (electrostatic) raised the
figure to approximately 9.2 to 10.0. The magnetic core storage
unit has a slight edge and has raised the figure to a present
ceiling of the order of 10.5. Tremendous strides will have to
be made to push this frontier further. A figure of 12 would
correspond to having a megabit of storage at ones disposal with a
random or average access of one microsecond to any single
location.
Table VIII is a tabulation of the various log10 Capacity/Access
figures for various media and computing systems.
E. - F. INPUT-OUTPUT
Previous discussions on arithmetic units and storage devices
have shown the great strides that have been made in these fields
ditring the past several years. Arithmetic operation and storage
access times have decreased and storage capacity increased. Yet,
the commication link between the person and the machine is
relatively unimproved. Paper tape and cards, inherently bulky,
are prevalent and relatively slow. The main convenience afforded
by cards, particularly in couerical systems, is their capability
of handling an item of information, such as data on an insurance
policy, a payrofl line, a stock item, a set of corresponding test
data, etc. There is no doubt tbat punching cards is slow. Paper
tape perf orators are also relatively slow in the sense that the
data to be punched can be available at a rate faster than they
can perforate, although high speed perforators are being developed
and are finding application. Advances in the field of
photoelectric
TABLE VIII
LOG10 CAPACITY/ACCESS OF HIGH SPEED STORAGE UNITS
LOG10 CAPACITY/ACCESS SYSTEM
-------------------------------- ------
10.87 MANIAC-Il
10.74 UNI-SCI (ERA llOSA)
10.39 IBM-704
10.20 NORC
10.04 JOHNNIAC
9.85 IBM-705
9.78 IBM-701
9.66 NAFEC
9.61 PER MARK-Il
9.61 WHIRLWIND-I
9.55 MANIAC
9.49 UNI-SCI (ERA lOS)
9.35 ILLIAC
9.55 ORDVAC
9.51 ORACLE
9.21 lAS
9.14 SEAC
9.09 RCA BIZM&C
9.04 UNIVAC-Il
8.77 SWAC
8.52 EDVAC
8.1S RAYI2AC
8.08 NIDAC
8.08 UNIVAC
8.05 TECHNITROL-180
8.03 DYSEAC
8.03 FLAC
8.05 SEAC
7.40 IBM-650
7.50 ENIAC
6.80 IBM-608
6.11 ENIAC
4.84 TIM-Il
paper tape readers have brought about increased reading rates.
Keyboard input systems are useful only in small systems, since one
cannot detain a large expensive machine through the use of a
manual input. The only exception to this is the use of keyboards
on large scale computing systems for the insertion of words manually
for test purposes or other special purposes.
In addition to paper tape and card readers and punches, many
systems utilize high speed line and page printers as a medium of
output and magnetic tape as a medium of input and output. Magnetic
tape output still requires a conversion from magnetic tape to
cards
or printed page for use by human beings. Magnetic tape and wire
input is relatively in the same position. The prevalence of
various
input-output media for the 84 computing systems described in this
report is shown in the following table:
Input-Output Medium Prevalence
Manual Keyboard (Input) 25
Typewriter (Flexowriter) 29
Paper Tape (Mechan) 29
Paper Tape (Photo-Input) 21
IBM Card 25
Magnetic Tape 29
Magnetic Wire 2
Analog 5
Page and Line Printer 19
Relay, Neon, CRT Display 5
One method for decreasing the tine spent `waiting for reading
and writing instructions to be carried out is to provide for con-
currency. The later machines have built-in circuitry for
permitting
reading and writing to take place during computations. Apparently
the only stipulation is that a given storage location does not
become involved in reading or `writing and some other instruction
at the same time.
Another method for reducing reading and writing time and to
avoid a large amount of lost time when a large amount of machine
reading and writing is necessary is to provide for reading and.
writing on a high speed device such as a magnetic tape or wire
unit and allow "conversion" to another medium to take place off
the machine at `leisure". Magnetic tape to card converters anci
inverters are becoming available as well as magnetic tape to
printed page converters. Paper tape and cards may sometimes be
considered as forms of storage. Magnetic tapes and printed pages
conserve more space than paper tape and cards. In addition it
may be necessary to have one computing system communicate with.
another. For these reasons, input-output media conversion is becoming
quite prevalent and conversion equipment is rapidly becoming
available. Since input-output schemes are so many and varied,
a complete treatment would be a subject for a report devoted
exclusively to input-output. A further study of input-output
devices may be made by examining the Input and Output sections of
the 84 computing systems described in this report. Information
on the characteristics of various input-output media is covered
in these systems descriptions. Various printing, punching, re-
cording, plotting and reading rates are given along with a
myriad of special features and remarks.
Many computing systems utilize several input- output media.
This has come about in an endeavor to provide compatibility
between systems, to increase input-output reliability and speed
through the use of more advanced techniques, and to satisfy
certain needs for output in different forms for various users.
The following tabulation shows an approximate count of the number
of systems utilizing one or more different input-output media for
the 84 systems described, in this report.
Number of Media Used Number of Systems
One 29
Two 52
Three or More 25
------ --
Total 84
G. CIRCUIT ELEMENTS
There are many impressions `which come to mind when one
examines such things as tube and crystal diode counts in a large
scale computing system. There is a tendency to visualize a
large, sprawling system when the tube count is high. There may
be large tube-changing programs based on experience in effect
on these large systems. Failure rates, preventive maintenance
techniques, tube life probeJans, design limitations and tube
specifications must all be considered on a regular systematic
basis when the tube count is high. Tube count may yield an
approximate estimation of some of the probelus that my be en-
countered in the operation of the system. Table IX shows the
approximate number of tubes utilized in some of the computing
sytems described in this report.
It is possible that the servicing of a large electronic
computing system can be materially simplified by reducing the
number of tube types in the system. Standards for tube testing
need apply only to a limited number of tube types and tube
checking
can be further systematized due to a reduced number of test
variations. Of course, a test specification or test criterion
must be established for the most severe application for which the
particular tube type will be applied. A severe or special circuit
TABLE IX
TUBE QUANTITIES IN VARIOUS COMPUTING SYSTEMS
QUANTITY SYSTEM QUANTITY SYSTEM
-------- ------ -------- ------
28,759 RCA BIZMAC 1,424 SEAC
17,468 ENlAC 1,400 CARAC
10,000 IBM-702 1,500 CAIJDIC
9,800 NORC 1,500 PEGASUS
7,200 WEIRLWIND-I 1,250 1314-604
6,100 IBM-705 1,224 DYSEAC
5,400 UNIVAC 1,050 FLAC
5,200 RAYDAC 1,000 CIRCLE
5,000 ADEC 900 MIDAC
5,000 J0HNNIAC 800 MONROBOT-IlI
4,700 UNI-SCI (ERA 1103) 700 BENDIX-D12
4,400 UNIVAC-Il 700 MINIAC
4,200 UNI-SCI (ERA llOSA) 600 MODAC-404
4,100 LOG 600 MODAC-410
4,000 IBM-701 600 ORDFIAC
4,000 ORACLE 600 WEDILOG
5,565 EDVAC 500 WHITESAC
5,500 FER MARK-I 481 HUGHES AAC MOD-III
5,000 IAS 450 BENDIX-G15
5,000 UDEC-I 450 ELECOM-125
5,000 UDEC-lI 425 NCR-CRC-102D
2,799 ILLIAC 400 NCR-CRC-102A
2,791 ORDVAC 400 BAR-COL DEC DIG
2,700 UNI-SCI (ERA ll02) 400 ELECOM-120A
2,695 UNI-SCI (ERA ll0l) 400 MELLON INST DIG
2,600 SWAC 550 NCR-303
2,600 TECH-180 290 ELECOM-125 FP
2,584 1314-607 280 ALWAC-III
2,550 MANIAC-II 265 READIX
2,400 MANIAC 240 ELECOM-lO0
2,000 IBM-650 180 TIM-Il
2,000 NAREC 165 BUR-ElOl
1,800 WISC 160 ELECOM-50
1,600 HAL RAY BROWN 140 MAGNEFILE-D
1,500 DATATRON 130 MAGNEFILE-B
1,500 FER MARK-Il ll2 OLIVETTI-GBM
1,500 IBM-CPC 100 LGP-30
1,500 PENNSTAC
requirement may well be better served through the use of another
tube type. This, then increases the number of tube types.
Normally,
it is possible to select a type of tube for a group of duties. In
a given system, for example, a certain type is selected for
driving,
for voltage amplification, for flip-flop circuits, normally "on"
or
"off" conditions, etc. This establishes a number of tube types for
a given system and any modification of the system will include
this
"tube type" complement. To show the prevalence of the number of
different types of tubes in various computers described in this
report, refer to Table X.
The question of crystal diode reliability, diode testing
techniq.u.es, and diode logical construction, such as individual
clamps versus wired plug-in units, printed circuits, etc. cannot
readily be resolved. The quantity of diodes in a given computing
system may be indicative of the nature of the servicing problem,
but only `when the failure rates, life and circuit demands placed
upon the diode are known. To some extent, malfunctions due to
diodes can be aggravated at elevated temperatures. The extent
of crystal diode use in the computing systems described in this
report is shown in Table XI.
Go to Next