AI summary by TwinMind
Why the UK's IBM Failed
Asianometry
34m 48s
Origins of Punch Card Technology (1880s-1920s)
- Hermann Hollerith invented electromechanical tabulating machine for 1890 U.S. census
- Founded Tabulating Machine Company (TMC), later sold to Computing Tabulating Recording Company (CTR)
- CTR renamed to IBM in 1924
- British Tabulating Machine Company (BTM) founded 1907
- R.P. Porter and Sir Raleigh Philpotts licensed Hollerith technology
- Exclusive license throughout British Empire (except Canada) for 25% of revenue
- First automatic tabulator 1908: processed 150 cards/minute
- Used for 1911 England/Wales census, popularizing technology
BTM vs IBM Performance Gap (1920s-1949)
- Major competitor: Powers Samas (French-American conglomerate)
- Performance disparity: IBM's American territory only 2x larger than BTM's but generated 20x revenue
- IBM dominated U.S. with 90% market share; British Empire split evenly between Powers Samas and BTM
- Thomas Watson blamed BTM's "low-key, typically British" sales approach
- BTM blamed onerous 25% royalty rate as "millstone around their neck"
- 1937 tension: Watson offered to buy BTM stock, Philpotts refused
World War II Impact and Transition (1939-1949)
- BTM had ~1,200 employees at war outbreak
- Both firms stopped R&D, converted to military production (aircraft frames, gun aimers)
- Post-war: rebuilt sales organizations with demobilized troops
- Technology shift: stored program computers emerged during war
- Early computers: house-sized, cost over $1 million
- Industry approach: "evolution not revolution" - gradual adoption into existing products
IBM's Electronic Evolution
- IBM 603 electronic multiplier (September 1946): read 6,000 cards/hour
- IBM 604 electronic calculating punch (1948): added programming capability via vacuum tubes/plug board
- IBM 604 massive success: sold well beyond expectations, 400+ units still in use by 1974
BTM-IBM Divorce and British Computing Birth (1949)
- BTM and IBM mutually dissolved decades-long agreement
- BTM revenue: £5 million vs IBM's $400 million USD
- License benefits lost: no IBM products/R&D access, but 25% royalty eliminated (tripled operating profit)
- National Research Development Corporation (NRDC) founded 1949
- Lord Halsbury predicted punch cards would be museum pieces within 10 years
- Began pushing British companies into computing
Early British Computer Companies
- Ferranti: partnered with Manchester University, first computer installed February 1951 (4 months before UNIVAC)
- Elliott Brothers: second British firm with commercial computer (1953), focused on low-power niche
- English Electric: diversified manufacturer, worked with National Physical Laboratory
- BTM produced Hollerith Electronic Computer (HEC) with University of London
IBM's Computing Breakthrough (1955-1962)
- IBM 650 (1955): world's first mass-produced computer, first profitable computer
- Only 2 IBM computers shipped to British Isles in entire 1950s
- IBM 1401 (1956): replaced vacuum tubes with transistors, cheaper/faster/more reliable
- Complete ecosystem: hardware, peripherals, printers, mass storage, software
- Sold 10,000+ installations, became IBM's most important product by 1962
- Obsoleted all first-generation vacuum tube computers
British Industry Consolidation (1958-1964)
- BTM acquired Powers Samas (1959) → International Computers and Tabulators (ICT)
- 17,000 employees, digestion problems from merger
- ICT acquisitions: GEC computers (1961), EMI computers (1962), Ferranti computers (1963)
- Gained key product: Canadian-made Ferranti Packard 6000
- English Electric acquired: LEO computers (1963), Marconi (1964)
- Result: only two major players left - English Electric and ICT
Technology Gap and System 360 Challenge (1964)
- Software crisis emerged: software costs growing, sometimes exceeding hardware costs
- IBM System 360 (1964): $500 million investment, single scalable architecture
- 6 processors, 40 peripherals, replaced all previous IBM products
- Third generation computers, caught industry by surprise
- ICT response: reworked Ferranti Packard 6000 → ICT 1900 series (rushed to market)
- English Electric: licensed RCA architecture compatible with System 360 → System 4 (1965)
Government Intervention and Political Pressure (1964-1968)
- Harold Wilson government established Ministry of Technology (MinTech) 1964
- Concerns about American technology dominance
- GE purchased controlling share of France's Bull (1964)
- U.S. denied export licenses for computers to prevent French H-bomb development
- France's Plan Calcul: $100 million investment in national computer champion (ultimately failed)
- UK government pushed English Electric-ICT merger
- Initial resistance (1964), companies felt strong
- Second attempt (1966): offered £30 million dowry
- 1968 UK economic crisis: dowry reduced to £13.5 million
ICL Formation and Challenges (1968-1970s)
- Merger finalized July 1968: International Computers Limited (ICL)
- £100 million annual revenue, 34,000 employees
- World's largest non-American computer manufacturer
- Inherited incompatible product families: 1900 series (ICT), System 4 (English Electric)
- Political pressure prevented phasing out technically inferior 1900 series
- Focused on high-end mainframes, missed minicomputer boom
- 1968 computer industry recession eliminated GE and RCA
- Conservative government (1970): anti-interventionist stance, merged MinTech into Department of Tech and Industry
Financial Crisis and Government Rescue (1970-1981)
- ICL received £40 million government loan (1971) after laying off 3,000 workers
- R&D costs escalating: 50% increase from 1974-1978
- Two major cost drivers:
- Software development: over 1/3 of entire R&D budget
- Accelerating semiconductor production cycles
- 1980 recession: £50 million projected loss, exchange rate hit 2.35 USD/GBP
- Thatcher government: £200 million loan guarantee, demanded management replacement
- 10,000 jobs cut from 1980-1985
Fujitsu Partnership and Decline (1981-2000)
- ICL sought semiconductor partnership
- American companies offered 3,000 logic gates/chip
- Fujitsu offered 5,000 gates (later revealed 8,000), plus 5-year roadmap visibility
- 1981 technology transfer deal: ICL peripherals for Fujitsu semiconductors
- Series 39 computers launched 1985, but mainframe market already dinosaur
- Standard Telephone and Cables (STC) acquired ICL 1984
- ICL evolved into assembler/distributor: 50% gross margins but no innovation
- Fujitsu acquired ICL 1990, shut down brand 2000
Lessons from ICL's Failure
- Lack of technical primacy: always following IBM and Japanese competitors
- Industry splintering: predicted vertical integration didn't materialize
- Specialized companies (like Microsoft) raised stakes beyond what combines could match
- Hollowness of national prestige: tried to beat IBM at its strongest point (mainframes)
- Ran into same market forces that caused IBM's 1990s decline
