I.
Historical Narrative The 1950s
Table of Contents
The Beginnings of the EDP Business: The 1940s 1.
2. 3.
Evolving EDP Technology . • •
Potential Early Entrants IntoEDP • Sources of Uncertainty About the Commercial Possibilities of EDP •
II. First Attempts to Commercialize computer Systems: The Early 1950s
4.
5.
6.
Remington Rand's Entry a. Eckert-Mauchly
b. Engineering Re·search Associates (ERA) c. The Leadership Position of the
Merged Companies • • • • • • IBM's Early EDP Involvement
a. The Defen~e_"Calculator or IBH 701 b. The IBH 650
c. The IBH 702
other Early Entrants Marketing Computers Commercially • • . .
[7/23/80]
Paqe
----1
1
7
.J..3 13 13 17
22-24 33 39
44-46 7. Customer Ignorance, Uncertainty, and Fear . . . 49 8. Expanding the Market for EDP Products
and Services . . . • . • . a. Short-Term Leases
b. Customer Support
53
53
III. IBH's Commitment to the EDP Business:
the mid~1950s
9. SAGE: IBM's Role and the Effect on IBM's Position in the EDP Indus~ry 10. The IBM 704 'and 705
11: FORTRAN
12. [IBM Entry into EDP Business--1955] 13. The IEH 305 RMlAC
14.' !LM's 1956 consent Dec~~e 15. The IBH 709
16. [IBM's Decision to Concentrate on the BDP Business--1955-57] • • . 17. Remington/Sperry Rand,
2
69 69 80 86 90 93 98 102
106 107 a. Remington Rand Lacked Commitment to EDP 107
(i)
(ii) (i~i)
Inadequate Marketing
Lack of Product Developments Loss of Key Employees. . . •
b. Conflict among Remington/Sperry Rand's
113 117
120
Divisions • • • . • . • • • . . • • 124 18. Other Companies
IV. The Second Generation
19. STRETCH
20. IBN's Second Generation Commercially Available Computer Systems . . • a. IBM's 7000 Series
b. IBM's 1400 Computer Series
c. The IBM 1620 Computer
3
Page
d. IBM's Second Generation Disk Drives
·
· ·
·
15221. Sperry Rand
. .
. ·
·
·
·
·
· · · ·
·
· ·
· · ·
157a. LARC (the "Livermore Advanced Research Computer" )
· ·
· · ·
· · ·
159b.
The "Solid State Computer"·
-
· · · ·
162c. 1105
. . · · ·
· ·
·
·
·
· ·
165d. New Large Scale Computers and Related Peripherals
·
· · ·
165(i) [UNIVAC III]
·
·
· ·
·
·
· ·
·
169(ii) UNIVAC 1107
· · · · ·
· · · ·
·
·
169(iii) UNIVAC 490
·
·
·
·
·
· ·
·
· · · ·
-
" 171 e. lvlilitary Projects ._·
·
· ·
-
·
· ·
· ·
172f. Gemini Committee
·
· · · ·
·
· ·
·
1"'" i ..) 22. Other Companies·
·
·
·
·
·
·
· ·
·
17423. American Telephone & Telegraph
·
· ·
·
·
·
17724. Raytheon/Honeywell
·
·
· ·
· ·
·
·
· ·
185,; 25. RCA
.
. . ·
.
. · · ·
· · · ·
· · ·
·
·
195a. RCA's Early Computer-Related Activities
·
:. 195 b. RCA Computer Developments 1956-1959·
·
·
·
197c. RCA's Computer Developments 1960-1963
·
·
·
20226. General Electric
·
·
·
·
·
·
·
· ·
20827. Electrodat<l
· · · ·
·
·
· ·
·
·
·
·
21728. Burroughs
·
· · · ·
· · ·
·
22229. National Cush Register
· ·
·
· · ·
2344
Page 31. Control Data Corporation (CDC)
. .
.
· ·
·
246 32. Technological Progress.
. .
·
·
·
25733. Reasons for IBH's EDP Success through
[7/22/80J
Historical Narrative The 1960s
Table of Contents
IV. IBM's Systern/360 and the 1960s 34 . IBM
a. b.
3000 Seri~s and th~ Spread Committee The Spread Repo"l'
(i) Price/Pe~ ..Jrn. logical LeadE:
'nd S/360 . . . . . ~ and Techno-a.4ip--Generally (ii) Systern/360 Component Technology
269 269
27]
278
279 282 (iii) Single Family for All Applications 290
(iv) System/360 Compatibility
(v) Emulation
(vi) Systern/360 Soft\..;are
(vii) System/360 Peripherals .
(viii) Standard Interface/Modularity Appendix: Examples of System/360 Uses 35. The System/360 Commitment . . • .
a. Preparation for Announcement
b. 360's Success and Impact on IBM.
36. Initial Competitive Responses to System/360
a. RCA
.
. . .
·
· ·
.
.
·
·
.
.
.
b. GE
·
· ·
· · .
c. CDC
.
. .
.
·
·
· . . ·
·
.
·
·
·
·
·
·
296 304 311 314 332 Al
341 357
367
376
d. Sperry Rand
.
.
.
-e. Burro~ghs
f. Honeywell
.
.
.
. . . .
.
. . . . .
g. SDS. .
.
. . .
.
37. IBM's Responses (1964-66)
.
.
.
.
.
.
.
.
.
a. Reduction of Extra Shift Usage Charges . b. l1C!l1ory Improvements . • ~
c. Tape Drive Improvements •.• . . d. Disk Drive Improvements •
e. Introduction of the Nodel 20
.
.
. . .
f. The Model 90 Program g. The 360/44
h. The 360/67
38. Educational Allowances
IBM's Educational Support Programs
39. IBM I S UnDundling
. .
. .. .
.
. . .
.
a. Introduction
.
.
.
.
.
. . . .
..
.
b. The Continued Demand For Bundlingin the 1960s • • • • • • • • • • • • c. IBN's Unbundling Announcement.
40. Sperry Rand/Univac
a. univac's Problems in 1964 • • b. The 1108
c. The Product Line Task Force .
2
382 383 383 383 385 385
.386
d. The 9000 Series • •
.
.
.
.
.
.
. .
.
.
e. Univac's Success in the Late 1960s 41. General Electric
. . .
. .
. . .
.
.
42.
a. The GE 400 Series • b. The GE 600 Series
c. Time Sharing
·
.
.
d. False Starts
·
.
.
e. The Management of GE's Computer'Operation . f.o GE IS Position in the Late Sixties • . • • •
h. ?he Ventures Task Force and the Decision to Disengage • • • • ~ • • • • • • • •
i. Did GE Lose Honey
· ·
RCA
.
.
.
. .
· ·
a. RCA 3301
. . . . . .
b. The Spectra 70 Series
·
·
(i) Compatibility with System/360 (ii). (iii)
Commercial and Scientific Ability of Spectra 70 Series . • • • • • Problems with Spectra 70 Series
·
· . .
·
. .
·
·
...
(iv) RCA Success with the Spectra 70 Series
c. RCA Computer Systems Division 1969-1971
d.
(i)
(ii) (iii)
Changes in Management Personnel and goals • • •
The RCA Series . . . • • • . • . Computer Systems Division's Prob1ems--Early 1970s • • • RCA's Decision To Sell Its Computer Business To Sperry Rand . . • . • •
Page 482 486 488 490 493 505 512 516 523
533 543 547 547 551 552
... 558 560 · 575 578
578 583 595
43.
44.
e. After the Sale to 'Sperry Rand • •
f.
(i)
(ii)
RCA's ~ctivities • • • • •
Sperry Rand's Success with RCA's Computer Systems Division • • • Conclusion
HoneY'vel1 •
a. The 200 Series
b. Problems and Solutions
(i) Otlv~r Systems
(ii) r~ri~hs~als c. Marketing Practices
d. Product and Service Acquisitions and Expansion • • • •
Burroughs • •
4
615 615
616 616 619 619
626 626
628 630
632
638
a. Burroughs in 1964: Problems and Changes 638
(i) Reduction of Expenses
(ii) Increased Product Development b. Computer Development 1964-1969
(i) The 500 Systems Family • (ii) Smaller Computers
(iii) Peripherals
.
.
.
.
.
. .
.
c. Burroughs at the End of the 1960s • 45. National Cash Register
46. Control Data Corporation (CDC)
a. CDC's 6000 and 3000 Series Offering (1963-1969) • • • • • • • • • • •
(i) The 6000 Series
640
642 643 644 650
653
654
657
670
671
5
Page '(ii) The 3000 Series
·
·
·
·
·
·
·
·
·
677-b. CDC's Expansion into Commercial
Data Processing
· ·
·
· · · · ·
678 c. CDC's Expanding Peripheral Business·
·
· ·
682 d. Data Centers·
· ·
·
·
·
·
· · ·
·
·
·
684 e. CDC's Acquisitions (1963-1969)· · ·
·
686 f. Conclusion· ·
·
·
· ·
·
·
·
·
69047. SLS
·
..· · ·
·
.....
·
.. 091a. The SDS Entry Strategy
· · ·
· ·
·
693 b. The SDS 910·
·
·
·
·
·
·
·
·
· · · · ·
695 c. The Expansion of the SDS 900 Series·
·
·
·
696 d. The Sigma Series· ·
·
· · ·
·
703 e. The Merger·
·
·
· ·
·
·
·
·
·
71148. Digital Equipment Corporation
·
713a. PDP 1, 4 , 5 and 7
· · · · ·
·
· ·
·
·
·
718 b. PDP 6·
·
· ·
·
· ·
·
·
· · · ·
· ·
721,J
c. PDP 8
·
· ·
·
·
·
·
· ·
· ·
· ·
·
·
722d. PDP 10
· · ·
· ·
·
·
· ·
·
..
.
727e. PDP 15
· ·
·
· · ·
729f. PDP 11
· · ·
·
· ·
729g. Peripherals and Soft\vare
· ·
·
· · · ·
731h. Competition
· ·
· ·
·
·
7326
Page
50. The Emergence of IBI1 Plug-Compatible
Manufacturer (PC}l) Competition
· ·
·
·
750a. From OEN to PCH
·
·
· ·
753b. PCr.! Entrants
· ·
·
· ·
·
·
·
·
762(i) Telex
· · ·
·
762(ii) Ampex
.
·
·
· ·
·
· · ·
·
765(iii) Mernorex
· ·
·
· ·
·
·
· ·
767(iv) ISS
·
·
·
·
·
·
..·
·
775(v) CalComp/Centu" 'ata Systems
· ·
·
776(vi) Sanders Assr .Lat_, - Inc •
·
·
· · ·
·
·
777c. PCB Price Competition and Success
· ·
·
78051. Leasing Companies
·
·
· ·
· · · ·
·
·
·
797a. An OvervieT,v of Leasing Company Operations
·
797 (i) The InvesL'"TIent Tax Credit·
·
·
·
800(ii) Harketing Costs
·
· ·
·
·
801(iii) Capital Availability and Cost
·
· ·
·
803b. The History of Leasing Companies in the 1960s
.
.
·
·
· · · ·
·
·
·
·
·
·
·
805(i) Leasing Company Growth
·
·
· ·
·
·
807(ii) The Emergence of Netv Challenges
· · ·
814(iii) Diversification
·
·
·
818c. Some Individual Companies
·
·
· · ·
819(i) Greyhound
·
·
·
·
·
·
·
·
·
·
·
·
819(ii) Boothe Computer Corporation
·
·
· ·
·
8217
Page d. ° The Effects of Leasing Companies on IBM 826
52. Service Bureaus -• • • •
.
. .
a • Entry and Gro~vth
.
. .
.
.
b. Time Sharing and the "Computer Utility" c. Competition • •53. Software Companies
54. The Role of the Federal Government 55. Planning for New Products • •
Introduction
Tape Drive Development: The 2420 and
3420 (Aspen)
·
· .
·
·
·
·
· ·
c. Disk Drive Developments: The 330 (Herlin) I
2319 and 3340 nvinchester)
·
·
·
·
· ·
(i) The 3330 . (Herlin)•.
· · ·
(ii) The 2319 and 3340 n';inchester)d. Ne,v Processor Planning eNS and System/3) (i) Monolithic Logic and Hemory
·
· ·
(ii) System/3·
·
·
· · · ·
(iii) Virtual Hemory
·
· · ·
·
·
· · · · ·
·
(iv) NS Prices
·
.
·
· ·
·
·
·
e. Conclusion· · ·
·
·
· ·
· ·
·
56. Grot,oJth of the EDP Industrya. Increase in the Nwnber of Users of Computers b. Expansion by Existing Computer Users
· · ·
c. Explosion of Ne\v Applica tions of Computers. d. Improved Price/Performance and Ease of Use. 57. Conclusion
.
.
. ·
·
. ·
·
· ·
·
·
· ·
·
· ·
·
831 831 835 840 851 866 878 878
885
898 898 902 907 907
911
913 918 922 923 926 928 929 937 944
[8/1/80]
Historical Narrative The 1970s
Table of Contents
58. An Overview • . • • • VI. Introduction of System/370
59. Initial Systern/370 Announcements
a. Competitive Pressures: 1970 •
b. IBM's Initial Systern/370 Announcements
(i) (ii)
(iii) (iv)
Model 155 and 165 Processors IBH 330/3830 Disk Subsystem IBM 3211 Printer . •
IBH System/370 Nadel 145 Processor, the IFA and the 2319 Disk • . . . . . (v) IBN 3420 Tape Subsystem
953
956 956 961 963 965 967
968 971 (vi) IBM 2319B and Disk Price Reductions 972 (vii) System/370 Model 135 Processor
(viii) IBM 3270 Termina,J. Subsystem (ix) IBH's Fixed Term Plan
60. Product and Pricing Actions of Competitors 61. Computer Systems Hanufacturers
a. Burroughs. • •
. .
.
.
. .
. .
.
. . .
b. Control Datac. Digital Equipment • d. HoneYVlel1. . . .
e. NCR. . • • • f. Sperry Rand • •
62. IBM plug-compatible peripheral equipment
manufacturers . • •
a. Telex.
b. Storage Technology
c. Memorex
.
.
.
. . . .
.
.
. . .
. .
. . .
d. Ad-vaIlGed I1.;:.;mory Sys terns (AMS)63. Leasing Companies .
64. IBI-! System/370 Announcements: 1972-1974 . .
a. IBH 3705 Communications Cont;roller b. The August 1972 Announcement
(i) 158/168 Processors: FET Memory
·
·
·
(ii) Virtual Hemory Operating Systems·
(iii) The New Attachment; Strategy
·
·
·
c. IBM System/370 Models 125 and 115
.
d. Term Lease Plan
.
. .
. .
.
.
·
· ·
e. 1973 Disk and Tape Subsystem Announcements(i) "Birch"
(ii) "Winchester" (iii) "Iceberg"
f. IBH 3600 Terminal Subsystems g. System/3 Model 15
h. IBM 3850 Mass Storage Subsystem . VII. Expansion and Entrv of Competitors
65. Introduction • • • • e. • • • • • • • • • •
2
995 1000
66. Growth of Competitors: 1974-1980 67. 1960s Competitors.
a. Burroughs • •
.
.. .
.
b. Control Data Corporation(i) Peripherals
.
·
·
. .
·
·
(ii) Data Services.
.
· ·
·
·
(i:'i) New} Comput . .a.c L~:..z ..
· ·
. .
·
·
c. Di~ital Equipment Corporation (DEC) d. Honeywell".(i) (ii)
Xerox
. . .
.
.,...
.
.
CDC/Honeywell Joint Venture (iii) ell/Honeywell-Bull.
(iv) 'GE • • • (v) Incoterm. (vi) Synertek. (vii) Spectronics
(viii~ GE Joint Venture e. Hewlett-Packard . •
f. Memorex. g.
h.
NCR • .
Sperry Rand • (i)
(ii)
Product Actions Acquisitions . i. Storage Technology
. .
.
.
·
·
·
· ·
·
· · ·
·
·
·
3
(i) (ii)
"Hinicomputer Peripherals" • Communications
(iii) Components. •
j. Xerox Corporation (i) The Recession (ii)
( i i i )
(iv)
(v)
Product Obsolescence • • Manufacturing Problems . Organizational Changes . Competition.
68. Entry and Expansion of Newer Competitors 69. So-called "Minicomputer" Manufacturers
70.
a. Data General b.
c. d. e. f.
g.
Prime Computer, Inc. Perkin-Elmer (Interdata) Harris Corporation
Wang Laboratories, Inc. Tandem Computers, Inc.
Datapoint • . • .
.
.
.
. . . . .
If Plug-Compatible Processor" Manufacturers
a. Amdahl Corporation . . . b. Others . . • • • • . . · . . . .
(i) Control Data Corporation . (ii) Magnuson Systems Corporation (iii)
(iv)
Two Pi Co., Inc. . . . . . . . . . Nanodata Computer Corp. . . . . .
4
(v)
(vi)
Paradyne Corporation Others •
. . .
11951198 71. Semiconductor Manufacturers
. .
. . .
.
1199 a. Intel Corporation . • • ..
. .
.
.
.
1200b. National Semiconductor Corporation 1203
72. Communications Firms 1209
73.
a. AT&T
(.i ) The Dataspeed 40 Terminals .
,.~ 'J
..&..~.L..:..
1213
( -. , ~ Dimension Private Branch Exchange (PBX) 1214
(iii) Transaction Netw<?rk Service (TNS) 1215
(iv) Advanced Communications Service (ACS) 1216
(v) Soft'\vare 1217
(vi) ESS and Other switching Systems 1217
b. Northern Telecom, Limited • 1218
c. FCC Computer Inquiry II • • 1223
Other Compe t i tors . • •. 1226
a. Software Companies 1226
(i) Cincorn Systems, Inc . . 1228
(ii) System Development Corporation . 1231
(iii) Computer Sciences Corporation 1236
(a) Contract Services . • 1236
(b) Data Services . . 1238
b. Service Bureau/Time~Sharing Companies . 1241
(i) Automatic Data Processing, Inc. (ADP) 1241
74.
(a)
(b)
Hospital Corporation of America (HCA)
The Wurlitzer Company •
6
1245 1245
(iii) HcDonnell Douglas Automation Company. 1247
c. Exxon Corporation • . (i) Zilog, Inc.
(ii) Periphonics Corp.
(i~i) Ramtek Corp.
(iv) Vydec, Inc.
·
·
(v) Others
.
.
·
·
·
Foreign Competitors·
·
.-a. Japanese Manufacturers b. European Manufacturers
(i) (ii) (iii)
Siemens AG • ICL, Limited • Nixdorf, AG
·
·
·
1250
· ·
·
1250· ·
· · ·
·
· ·
·
·
1254· ·
·
1255·
· · ·
1256·
·
· ·
·
·
·
·
1257·
· · ·
12581259
..
.
. . .
.
. . . .
12671267 1269 1271-VIII. IBM: 1975-1980 1274 75. IBH Product and Pricing Actions
·
·
· ·
1274a. System/32
. .
.
· . · · .
.
· ·
·
·
·
·
·
·
1287b. IBLvl 3800 Printing Subsystem
·
·
·
·
·
·
1292c. IBH 3350/3344 Disk Systems
· ·
·
· ·
·
· ·
1295d. IBH 5100 Computer System
.
·
· · · · ·
·
·
1301 e. Attached Processors: 168 and 158 1304 f. IBM System/370 138 and 148 Processors . 1307h~ Memory Price Reductions and the 303X Processors • • • •
i. IBH Announcements: October 1978-January 1979 • • • • • . . • .
(i) November 1978
(ii) December 1978
(iii) January 1979 • • • (iv) February 1979
(v) March 1979 . (vi) April 1979
(vii) May and June 1979 (viii) July-September 1979
(ix) October 1979 .
7
1313
1317 1328 1329 1329 1330 1331 1333 1335 1336 1336 IX. Customer Alternatives in the 1970s: Some Examples 1338
76. Introduction 77. Chemical Bank.
a. Overview of Chemical's EDP Installation: 1978 • . • .
b. Chemical's major EDP procurements
(i)
(ii) (iii) (iv)
(v)
(vi) (vii)
Branch t.erminal sys tern . Fund's Trans fer . . '. . Trust and I~vestment
Foreign Exchange Trading • . Legal Department
Fixed Asset Accounting . . Bond Trading . .
1338 1341,
78.
79.
(viii) Personnel
. . . . . .
.
.
.
.
. .
.
.
(ix) Bank investment and others •American Airlines
a. The Gro\vth of American Airlines'
Computer System • • • • • • • • • • • . (i) American Has Upgraded the Central
Processing units in its System
(ii) American Has 1)':"s ~rlDu·~cd t·.I.uch of its Processing • . . . . . (iii) American Has Expanded Its Storage
Capability . • . • . • • • • • • •
b. The Growth in the Number of American's EDP Suppliers . . • . • • • • • • • c. Implementation of American's Passenger
Reservations Applicat~ons . . • . • . . d. Implementation of Ot~er Applications on
American's EDP System. • . . . . . (i) Flight Planning
(ii) "PRAS If and "MOUS II
(iii) "Sky Chefs"
(iv) Americana Hotel Reservations . (v) Message switching
Union Carbide . • •
a. union Carbide's Configuration.
b. union Carbide's System Configuration: 1978 c. The Implementation of Union Carbide's
Order Entry Tasks . . • . . • • . • . • ~ . d. Other Current Applications at Union Carbide
8
Page 1370 1371 1374
1374
1375 1376
1377
1379
1380
1388 1388 1391 1395 1396 1397 1399 1399 1402
9
Paae
_ _ J _
e. The Battery Products Division's Request for Proposal: A Case study of Alternatives
to Users
. .
.
. ..
.
. ·
..
·
· ·
·
· ·
1412(i) Configuration options: the nRFP" 1413 (a) Centralized with a host computer at Tarryto~vn
· · ·
· ·
·
·
· · ·
1413(b) Centralized with a host computer at BPD headquarters
· · · · ·
· ·
1414(c' Dis tr :1.')'.l.1:8('. '1;'71 tl". :.. r.~c s-l: ':=orrlpu t.2:' . at TarrytoHn
·
· · ·
·
· ·
·
1414(d) Distributed with a host at Rocky River
· ·
·
·
·
· ·
1415(e) Stand-alone. ~ystems
· ·
·
·
·
· ·
1415(f) Service bureau host
· ·
· ·
·
1415(ii) The competitors' re?ponses
·
· · · · ·
1416(a) Burroughs
· ·
·
·
· ·
·
·
· ·
1416(b) Comserv
·
·
·
·
·
·
· ·
· ·
·
1417(c) Honeywell
· · ·
·
·
· · ·
· · · ·
1418Cd) IBN
·
·
·
·
·
·
· ·
·
·
·
·
·
1422(e) Hartin Marietta Data Systems 1423 (f) Univac
·
·
·
·
·
· ·
·
· · ·
1426(iii) Union Carbide's selection
·
·
· · ·
·
142780. Southern Railway
·
·
·
· ·
·
· · · ·
1429a. Introduction
·
·
·
·
·
·
·
·
·
1429b. Design alternatives available to users generally
.
. · · ·
·
·
·
·
·
·
·
·
·
·
1430(i) Southern Pacific
·
·
·
1432c.
(iii) Illinois Central Gulf •
(iv) Santa Fe
(v) The British Railroad
(vi) The French Railroad's.
(vii) Canadian National and Canadian
Pacific Railroads • • • . • •
(i) Missouri Pacific
(ii) Union Pacific
(iii) Conrail (Penn Central)
(iv) The German Railroad
Evolution of Southern Railway's
Computer System • • .• • • • • .
1433 1433 1433 1433
1433 1433 1434 1434 1434
1435
d. Southern Railway's "major" acquisitions 1437
(i) The 7040/44 Acquisition 1437
(ii) The System/360 Model. 30 Acquisition. 1439
(iii) The System/360 Model 65 and 50
Acquisition . • . • • .
(iv) The Burroughs TC-500 Acquisition
(v) The Univac DCT 1000 Acquisition .
1440 1442 1444
(vi) The Svstem/370 Model 158 Acquisition 1445
(vii) The Four Phase Acquisition 1447
(viii) The Data General Acquisition 1448
e. The Implementation of Southern's
Applications . . . • • • . • • .
"TIPS"
(i) Division of work between the central
1451
and remote sites • . . . . . 1451
(ii) The "TIPS" applications and the
."
"(iii) The implementation of particular TIPS functions at Southern Railway
(a) Yard Inventory
(b) Terminal Inventory (c) Haybilling
(iv) Implementation of TIPS applications at other railroads • • • . • . • •
(a) San'ta Fe.
~~) Seaboard Coastline 'Illinois Central Gulf
11
1454 1455 1456 1456
1457
1458 1458 (d) Missouri Pacific and Union Pacific 1458
(e) Southern Pacific 81. General Motors
a. Overview of COIllputing at General Hotors • • b. Development of Computer Aided Design
for Automobiles at GMR . . .
c. The Competi tion Bet\'leen IBM and CDC at GHR for Its Automobile Design Application • • • d. Organization and Purpose of GMR's Computer
Science Depar~~ent at the Time of Hart's Testimony . • • . . •
e. Computer Applications Performed at Gi'-lR's Computer Science Department at the Time of Hart's Testimony • • . . • • . . . • • . . f. Computer Equipment Installed at GMR's
Computer Science Department at the Time of Hart's Testimony . • . . • . • • . . . • • g. GMR's Computer Procurement Procedure
1459 1460 1460
1462
1464
1467
1467
1468 1469 h. Cost Effectiveness of ,Computing Alternatives
i. Computing Alternatives Evaluated by the Computer Science Department . • • . • • j. Hart's View of the Performance of the
Computer Industry--Past, Present and Future 82. Firestone • . • • •
85.
a. Introduction
b. Firestone's EDP Installation: 1977 c. Alternatives Available to FirF?st0ne Federal Government
',... 1\. '. _. L.i
a.
b.
Account Studies .
"Large Systems Product Plans" . "Hinisystem Highlight Reporting--Account Profiles" • • . • . .
(i) First National City Bank (Citibank)
(ii)" J. C. Penney.
(iii) Massachusetts General Hospital
(iv) The Equitable Life Assurance Society of the United States . • • . . • . c. "Lo,,'; End Productivity Analysis"
(i) Southwest Bell Telephone
(ii) Dow Chemical • • (iii) Procter & Gamble .
(iv) Pepsi-Cola General Bottling
(v) Northwestern Mutual Life Insurance Co. (vi) Standard Oil of Indiana
Conclusion
12
1472 1476 1483 1483 1484 148:: 1492 1506 1506
1510 1512 1514 1515
HISTORICAL NARRATIVE VOLUME 1 OF 3
, ~ DEFENDANT'S .; EXHIBIT
J
11/
271
·1
I
I
Ij
1 1
1
I
2!
I
3 1
I
4
5 6
I
7
II
8II
9 10
111
12.11
I.ii
13 11
I,
14 !I 'I
I:
!! 15
I:
. 1i'
16 II iI
Ii
I17 I
I
18I
1920 21 22 23 24 25
II
Ii
'I
II
1\
HISTORICAL NAR~.TrvE STA~~NT
OF RICHARD B. MANCKE, FRANKLIN M. FISHER
AND J~~S W. McKIE
Introduction
The pages which follow represent our attempt to place the
record of this case into an historical perspective. We believe the
question whether IBM today possesses or at any time has possessed monopoly power in any relevant market can only be assessed by
reviewing the history of the EDP industry from its birth to the
present and in so far as the evidence permits into the future as
well. We therefore have prepared for the Court our analysis of the
major events in the life of the computer industry over the past 30
years, as refle~ted in the record of this case. We do not sugge~t
that we have s~a=ized for the Court i~ ~~is historical nar=ative
every fact or opinion which appears in the record of this case. We
have attempted, however, to set for~~ those events which appear to
us to be the most significant in ~~derstandi~g the development of
~~e industry, the position of IBM within the industry and the
reasons for the great success which Ia~ has had with its computer
products and services.
In order to avoid duplication and to expedite the massive job of culling through .the more than 100,000 pages of trial tran-script and the many thousands of exhibits and depositions which in their totality dwarf even the massive amount of transcript
avail-able, we have divided the task among the three of us. Dr. Mancke
prepared the initial sect~on of our historical analysis covering
-i-I
I
0-11
II
I
I
1
I
generally~~e
period from the beginning of the industry to the early2
3 4 5 6
7
8
9
10 11 12
I:
11
II
13 -II I'
,I
I' ,I
14 !\ I Ii
15
Ii
:1iI
16
II
!I 17il
18I
I1960s. Drs. Fisher and McKie then reviewed and discussed his draft
and concurred in the final product. In like manner, Drs. Fisher and
McKie prepared the second and third portions, respectively, of our historical analysis covering generally the period from the
develop-ment of IBM's System/360 in 1961 through the end of the 1960s and
then the 1970s.
In preparing our historical narrative, we were provided with assistance by IBM personnel assigned to work on the litigation
and by counsel for IBM in this case. Those people obtained from the
record (and other'sources) material when we requested it, checked our citations against the sources we utilized, put the citations
into a consistent format and proofread ~~d provided necessary
edit-ing, a~~i~istrative and clerical assistance.
~he historical narrative here presented represents the
product of our collaborat:on. We believe i t accurately and fairly
reflects the history of the computer i~dust=-I and ZEH's
participa-tion in it as reflected by the record and our ~,de=st~~di~g of t~e
19 record. Accordingly, we present i t to the Court as a part of our
20 test~ony.
21
22
23
24 25
-ii-1 I. TEE BEGI~I~GS. OF TEE EDP BUSI!reSS: THE 1940s
I
2 :1 1. Evolvinq EDP Technologv. Early research anc
.,
3 II
.,
I:1
development of computer technology was sponsored in
substan-4 :1
!I tial part by various branches of ~~e military and related
Ii
'1
5
~l
intelligence agencies who had extensive computational a~d5 :1 :1 data processing requirements. During Wo::-1d War I I anc
i
7 continuing thereafter, the United States gove::-noent was a
8
:i
9 :1
driving force in the EDP field, calling upon organizations
to build ever more advanced computer products. (DX 280; DX
10
iI
\1 3420A; DX 10283, pp. 6-7; DX 7528,' Mahoney, pp. 58-59;
ii 11 II II
.j Plaintiff's Acimissions, Set
IV, ~~~! 23.0, 48.0, 53.0, 20~. 0, I
I
12 ~ ! :j 221.0.)
13 ~ I Thus, the first large elect::-onic digital compu~er,*
;1
14 :1 :l the ENIAC, tvas developed during World War I! by a team of
'!
15
Ii
scientists/engineers, led by J. P::-esper Eckert and John N.;1
,,.'1 Mauchly, at the Universi ty of Pennsylvania's Hoore School
.0 'I
17 18
19
20
21
23
24
25
of Engineering ~~der contract with the United States A~y.
*
Digital computers are distinguished from analog cornpute::-sin that U(a] digital computer operates on discrete qua~tities'
and essentially counts", whereas "(a]n analog computer operates
in analogy with some physical phenomenon". (Fernbacn, Tr. 437.)
That is, an analog computer "solves problems by translating
physical conditions .such as flow, temperature, pressure, angular position, or voltage into =elated mechanical or electrical
quantities and uses mechanical or electrical equivalent ci=cuits as an analog for the physical phenomenon being investigatec.
!~ general i~ is a computer which uses an analog fo~ eac~
variable and produces analogs as output. Thus, a~ analog ccmpute=
measures conti~uouslv whereas a di=i~al com-;:,uter counts d:'sc:=etelv."
(DX 5202, p. 263; se~ also Beard,
Tr.
10195; JX 1, pp. 8, 39; •DX 4992, pp. 5-7; DX 5126, ~~.i-a.)
-1-'I
II :1 Ii !. 11 il 11
1 1\ !I
2 i\
:I
3
II
411
.15
il
Ii
6 :1 :1
ij
7 I. il
'I 11
8
II
9
II
Ii
10 it
1.1
11
iI
il
12 !!
'i I:
13 ~: !t
~ !
14 :i
i;
',1
.
,
~w ~I
:,
J.
.6 :i !: ~ ;
; ~
17 ii
~
I
18 Ii
19 I!
il
20 H 21
II
P II
22 :!
I!
II
ii
23 !I
:I 24
iI
'i
25 ::
il , ~
Ii
il " .1 I: :1 'I 1/ II:i
(Fernbach, Tr. 438-40; Eckert, Tr. 730-32; PX 1, p. 2; OX 5476, p. 26; OX 5423, Smagorinsky, pp. 8-9; OX 7532, Parten, o.
11; Plaintiff's Ad.'"I1issions, Set II,
,r
800.0.) ENIACwas designed to be used in calculating trajectories for
field artillery and bombing tables for ~~e
u.s.
~IrnyBallistics Research Laboratory at the Aberdeen Provinq
Ground. It "was developed specifically for ~~e ouroose of
generating firing tables. That was the original ournose
because, prior to that time • • .t.;'ey had a large number
of mathematicians who had to sit in rooms wit.~ desk
calcu-1ators, numerically integrating trajectories, and the basic
reason for developing the digital computer in the ~irst
place was to speed up ~~e process of ntl.'"Tterical inteqration."
(OX 7532, Parten, pp. 11-12.)*
The ENIAC was a physically enormous machine
(measuring
lod
feet long, 10 feet high and 3 feet wide, andcontaining about 18,000 vacuum tubes) and was described as
"one of the most complicated devices in the world". (Eckert,
Tr. 729, 77li Plaintiff's Admissions, Set II, ~r 800.2.)
Indeed, i t was so complicated that Or. Enrico Fermi reoortedly
*
ENIAC was also used to perform calculations for theAtomic Energy Commission at Los Alamos and to develoo and test models for "short-range (weather] prediction for the
Terrestrial Atmosphere". (DX 5423, Srnagorinsky, 00. 8-9: see
Eckert, Tr. 744-45'; Metropolis, Tr. 1133-34; Plaintif:'s
Admi s s ion s, Set I I, ~I ~I 5 57 . 4, 8 0
o.
6. )
-2-:
i
I
I • i I I
I
I I ,I
I
II
!I
II
II
II
II
11
II
II
1 \1 !I
2
!i
il
"doubted if the machine would run =or more th.an five
minutes at a time". (Eckert, Tr. 771.) In fact, when ~~e
3
i
IENIAC became operational in 1946, i t broke down only about
4
I, once a day. (Eckert, Tr~ 770.)
5
ii
II
6 ~
I
'I
7
II
I,!I
The ENIAC differed from prior computational
machines in that prior machines had all been
electromechani-cal--that is, ~~ey performed arithmetical calculations by
II
8 !I
II
9
using electricity to close mechanical relays. (Fernbach,
Tr. 438.) ENIAC's use of vacuum tubes rather than
electro-10
I
,i mechanical relays allowed i t to be' -faster than its
elect='o-11 i I
I mechanical predecessors by "at 'least a fac'tor of a hunc.=ed
I
12
!
j, and probably 500.". (Eckert, Tr. 758; see Fernbach,"
13
l!
:i T=. 439.) Nitn the ENIAC, it: TNas possible to 'gerform a wice1~ range of previously impracticable or impossible calculations.
;
,I ,
.
"
15 .,
:1
(Plainti== 's Ac.missions, Set ! ! , ff 800.13.)
~ ,.
"
lO T~e ENIAC had to be ?rogr~~ed by setting
switcnes--.
:..;-
anc whenever t~e ~=og=~~ needed to be changed, ~~e switches,18 numbering in the t~cusands, all had to be reset by hand.
19 (Eckert, Tr. 778; Metropolis, Tr. 1141-44; DX 5423,
II
20
"
'I:1 Smagorinsky, pp. 8-9; Plaintiff's Admissions, Set II,
:1
21 II
'. ~I~I 557.5, 800.7-.11.) This limitation was removed by· t.~e
\,
22 !: next major step forward in computing--the development of
" \I
23 II :1 electronic stored program digital computers. (Eckert, Tr.
:1
'.
24 "
:1 776-80; H. Brown, Tr. 82962; Plaintiff's Admissions, Set II,
ii
25 ~ i ~I 80 2. 4. )
!!
.!
II
-3-;;
I,
II
!l
I;
Ii
1
2
!\
II
II
il
I
I
i
3; i
41
!i
_ II
:l ;6 11
!, il
7 )1 :1
8
9 1
!I
10
!!
I'
11
!l
:1 II 12 H 13 1\. '
:1
, .1
."'1' I
:1
:
~ 15 :1'I 16
ii
IIa
17 II
:1
18 :1 :: 19
:i
:,i!
20 'I
q
~ f
'.
21 ii II ':
ii 22 I. \1
il
II
23 .!
11
24 ;1 : ~
I!
25 li
!n 1944, while the ENIAC was still under construc-tion, a group of people located at the Moore School, includ-ing Dr. Herman Goldstine, J. Presper Eckert, Dr. John. Mauchly, Dr. Arthur W. Burks, Adele Goldstine, and, after August 1944, Dr. John von Neumann, began to meet regularly to develop the conceptual design of an internally modifiable stored program
digital computer that became known as the EDVAC. (Eckert, Tr.
780-81; PX 5657, p. 2; Plaintiff's Admissions, Set II,
CJ 802.0-.1.) The "stored program" concept was based on
~~e realization that computer 'instructions could be
repre-sented as numbers and co~ld be stored in memory wi~ other
numbers, provided there was a way to identify t..~em as
i:lst=uc-tions • (Plaintiff I s Admissions, Set !I, ~r 802.4.) The
concept of "internal program modification" recognized that inst=uctions stored in memory could be handled and modified
arithmetically in the same way as other numbers stored in
memory.
*
(Plaintiff's Admissions, Set II, '1 802.5; see Hughes,*
A stored program is a series of instructions to thecom-putertelling it what·to do, and usually depends on either the results-previously achieved or the conditions existing
at the time tiie--,~omputations are made. (Plaintiff's
Admis-sions, Set IIi,'.."r 782.9.) In computers based on the "stored
program" concep't, instructions are stored within the machine
in the sarne fo~ as data. They are capable of being stored
anywhere in ~~e system, recalled from anywhere with the same
ease, or modified to the extent of ~~e capability of the
system. This capability of "computing" or processing parts
of the control program results in a far more flexible system
~~an had been known before. (Hughes, Tr. 33881, 33886-87.)
II
II,
i
I
I
:1
1 ~I
I,
2
!I
3 \1
:
I
41 .1 "
!
5
!
i
6i
il
7 !I
11
!I
8
!I
9 II
10
!
I
11
I
il 12 ~I
:!
:,
13 'I
li
14 'r :j
;!
15
I'
il
16
ii
I;
II
il
17
II
11 18 ii~ I
1I
19
il
20 II
11
2l
II
,I22 !I I,
Ii
,,~ I!
~;I
:1
24 jl
II
25 :\
"
II
II
il
Ij
I:
Tr. 33881; Hurd, Tr. 86405; Knaplund, Tr. 90461; DX 8988, ?p.
(Tr . 8 8 2 S 1) • )
The EDVAC's stored program concept was developed in detail in a series of papers written by, among others,
von Neumann and Goldstine. (Hurd, Tr. 86 327 - 28 ; OX 44, p. 5 i
Plaintiff's Admissions, Set II,
,r
802.2-.3.) These paperswere widely circulated after World War I I and were the
subject of extensive and intense discussion among a "very close fraternity of people" in universities, industry, and
government, working on designing an~ developing computers.
(OX 13526, Forrest, p. 66.) These persons
cornmuni-·cated actively wi~ each other about new circuits, new
devices and new co~puting machines by circulating technical
papers and attending symposia. (Hurd, Tr. 86327-28, 88206;
DX 5423, Smagorinsky, pp. 11-13; OX 13526, Forrest, p. 67.) In 1948 the Association for Computing Machinery was formed and quickly became the "premier technical society associated with computing". 'The ACM provided an organization (and an associated publication) in which "the scholarly and pioneer-ing work of computpioneer-ing could be laid down and distributed
into the society at large". (Perlis, Tr. 1853.)
In the late 1940s, following the initial scien-tific/technical discussion of the EDVAC stored program
concept, many ur.iversities, government-related laboratories,
and private firms began to design and develop stored progr~~
-5-2-3
i
I
• I
!
I
II :j
'I
II
II
I.
il
~ I
1
II
,
2 !
3
i r
,
4 ! "
5
ii
il6
~
:II
Ii
7 ,! "
!J
8 11
9
!I
:i :t
10 :j
;1
11 ;1 :1
, ~
12 :!
-,
13
:i
14, ... -j
--
"it "
16 II :1
;1
17 !1 il
~ I
18 n
!!
ii
19
.,
,: 1120 " -,
il
:'
21
ii
I, ,;22 d " i! I,
23 II
il
I(:, 24 :1
;j
"
25
il
Ii
11
II
ii
II
!;
:1
/, , .
computers, f;equently with government funding. A
list--which does not purport to be all-inclusive--of 21 nonorofit organizations designing and developing prototyoe stored
program digital computers in ~~is time frame is set !orth in
~~e footnote below.* Among the private firms engaqed in
designing and developing prototype electronic digital stored
program computers in ~~e late 1940s (often in connection with
military projects) were American Telephone and Teleqraoh,
Raytheon, Eckert-Mauchly Corporation, and Engineering Research
Associates. (Eckert, Tr. 773, 782; R. Bloch, Tr. 7566-70;
Hurd, Tr. 87662; DX 280.) The activities of these firms in
~~e la~e 1940s are discussed in some detail below and in the
company ?rofiles which fo~ a ?art of this testL~ony.
*
The Jniversity of Amsterdam; the University of Californiaat Berkeley (CALDIC) i the University of California at Los
Angeles (as operating agency) (SWAC); Cambridge Oniversi ty (EDSAC); the University of Frankfurt; Harvard University
(Mark III); the Universi ty of Illinois (ORDVAC, ILLIA.C); the
Institute for Advanced Study at Princeton {lAS ComDuter)i the University of Manchester; the University of Michiqan
(MIDAC); Massachusetts Institute of Technology (Whirlwind); the University of Pennsylvania (EDVAC): the University of
Rome; the University of Vienna; a university in ~weden; the
Federal High School in Zurich; the Los Alamos Scientific
Laboratory (MANIAC); Patrick Air Force Base (FLAC); the ~~
Corporation (JOHNIAC); the National Bureau of Standards (SEAC);
and the Naval Research Laboratory. (~, Hurd, Tr. 86324-26;
see also DX 5423, Smagorinsky, pP. ll~ Plaintiff's
Admis-sions. Set II, 11~ 558.0-.6; Plaintiff's Admissions,. Set IV,
'1 ~f 4 8 , 121.)
-6-13
14
2. Potential Early Entrants Into EDP. By the
early 1950s, the knowledge and resources necessary to build primitive computer systems were widely held and, therefore, many firms were well positioned to develop and supoly
com-puter systems. The most likely participants possessed one
or more of the following attributes:
(a) expertise in the relevant electronic and
.
electromechanical technology necessary to build com-ptlters (e.g., vacuum tubes, relays and transistors);
(b) experience in obtaining federal research and
development contracts (typically from ei~~er ~~e
military or intelligence aqe~cies) to design an= build
cne-of-a-kind data processing and/or control systems~
and
(c) expertise at selling products to ~~e =a~~er
small number of sophisticated organizations thoucht likely ever to purchase a computer system.
Examples of firms possessing these attributes included:
(a) Bendix, Boeing, Douglas, Hughes, North American
Aviation, Northrop, Raytheon, and Sperry who were high technology defense contractors with expertise in
designing and building sophisticated electronic control systems and were consumers of large amounts of comouta-tional power;
(b) General Electric, Westinghouse, RCA, and
-7-Philco who were large manufacturers of electrical equipment and had a broad base in the relevant tech-nologies, in addition to being potentially large data processing customers;
(c) American Telephone and Telegraph,
Inter-national Telephone and Telegraph, and General Teleohone and Electronics who had experience in manufacturing and consuming communication$ switching equipment; and
(d) Burroughs, Friden, IBM, Monroe, National Cash
Register, Remington Rand, Royal, and Underwood who produced calculators and/or business machines such as typewriters, unit record equipment, and accountinq machines.
AT&T, because of its early involvement in comDu~i~q
techniques, its huge size, Bell Labs' research caoabilities, and western Electric's experience as a defense contractor and large-scale producer of electronic and electz-omechanical products, was perhaps the best situated of all these companies.
In addition to the established firms listed above, there were a few recently formed, typically much smaller firms developing computer systems in the late 1940s and/or
early 1950s, often for the
u.s.
government. These includedEckert-Mauchly, Engineering Research Associates, Consolidated Engineering Corporation, Electronic Computer Corooration,
and Computer Research Corporation (a spin-off from Nor~~roo
-8-I
'I I.
II
II
II
1
!I
2
II
!
3!
! 141
Ii5 11
6
l'
'I:1 7 II
;1
!I
8 r
1
9 1
i
10 :,
r
11
!I
il
Aircraft Corporation). (Eckert, Tr. 805-08; Norris, Tr.
5599; Oelman, Tr. 6120-21; Hangen, Tr. 6262; McCollister, Tr. 10995-96; Withington, Tr. 55983.; Hurd, Tr. 88028; OX 280, p. 1; OX 12694.)
Finally, besides firms of the sort listed above, nonprofit, government-funded think tanks (such as the RAND Corporation) and the research affiliates of major universities
(such as Massachusetts Institute of Technology's Lincoln.
Laboratory) secured substantial federal funding from ~~e
military and intelligence agencies' -to build prototype
computer systems. (Crago, Tr.8596l-62, 86008-09; Hurd, Tr.
'2!!
86324-26, 88089-90, 88156, 88213-15.) In the formative~ ;! 13 14 15 16 17 18 19 20 21 22 23 24 25 :1 ;i I ,j
i ~
:j ! ~ :l .. :1 ;1 ii " i! ii ij ~ I !I " Ii .j I, I! :1 :1
~
i
!:
II
~ ! I! ii II I! ,. 1i II II II ;i .' d ij I: ., " "years of the EDP business, when, as we describe later,
nearly everyone believed that the size of ~~e total market
was severely lL~ted, these nonprofit organizations posed
substantial potential competition to their profit-making counterparts.
In sum, many firms were well-positioned to develop and supply computer systems and, typically with government funding, several had actually been developing compute= products.
-9-
I
I
1
II
i
I
i
i I
2' I 3 !
I
I
41
II
5
II
6 11
..
Ii7 !!
I' 8
II
911
10
II
II
11
l!
"I
12 !! 13 '4 ";
• I
:1
15
:j
h
16
ii
:i
17
!j
II
18
i!
19 :1
II
:1
20
il
2122
il
!Ii ,I ~
23 !I
I!
24 :1
II
Ii 25 [I
II j:
!I
II II
"
3. Sources of Uncertaintv About the Commercial
Possibilities of EDP. Though ~~e computer's potential for
performing large and complex calculations was wide"ly recog-nized by 1950, and though many companies had the knowledge and resources necessary to build computers, great uncertainty as to both the size of the potential market and the feasibility and costs of producing computer systems caused potential
entrants to be reluctant about actually investing substantial scientific, technical, production, marketing, managerial, and financial.resources to become "commercial suppliers of computer systems (as opposed to building prototype or
one-of-a-kind computers under contract for ~~e government). The
belief that there might not be a si"gni£icant market for
computer systa~s, which is described in more detail in ~~e
following sections, was deduced from the following premises:
(a) Only a few customers--primarily the militarj,
Weather Bureau, intelligence agencies, defense contrac-tors (especially airplane manufacturers)", the Atomic
Energy Commission and its subcontractors, and ~~e
Bureau of the Census--were thought to have computa-tional needs of sufficient magnitude and complexity to fully utilize a computer system as well as be able to afford such a system.
(b) Many of these potential customers, as well as
several major universities and nonprofit scientific
-10-I'
,III
I I :I
1 11
2
II
3
!I
4!I
51
I ! 61il
'I7
I.
i
8 1
9i
10
I
I 11
I
II
12
I,
\, ..l3 iI:! ,
14 15 ;j
II
16 ~ Ii I
~
i
17 I: II 18
i!
:1II
19 ;1 i! II
,I
20 :1 q
21 11
il
22 j:.1
23 11
li
24 il " I,
i! 25 II 1\ /I I! :1 II
"
'I
I.laboratories, were designing and building their own
computer systems.
(c) The first computer systems were physically
enormous, difficult to proqram, required complex
cir-cuit--y that, with the prevailing vacuum tube
technol-ogy, was prone to frequent failure, and were many
t~es more expensive than the most expensive electro-mechanical unit record equipment, business and
accounting machines then on the market.
(d) Few people had suf~icient training to be able to use a computer system. Most people skilled in
computer programming, utilization and maintenance were
t...~ose already employed by organizations t.."lat were
developing computers. Thus, to market ~~eir equipment on a commercial basis, ~~e m~~ufacturers t...~emselves
would have to provide users with most of ~~e program-ming, education ~~d support needed to operate the system •.
(e) Moreover, since the basic computer technology
was in the process of being developed, and engineering and production feasibility had not been demonstrated,
it was impossible to predict either costs or product
performance and reliability with any degree of
accu-racy.
Hence, though many large firms were well-positioned
