Electronic Computing: Crash Course Computer Science #2

Main ideas / lessons

• Growing world complexity created a need for faster computation

  • Early 20th-century systems (tabulating machines) helped automate bureaucratic/manual work, but the scale of human systems—population growth, wartime mobilization, global trade—created an explosion of data and administration.
  • This drove demand for more automation and computation.

• Electro-mechanical computing hit major performance limits

  • Machines became larger (room-sized) and hard to maintain, error-prone, and unreliable.
  • Harvard Mark I (completed 1944) illustrates both ambition and constraints:
    • 765,000 components, 3 million connections, 500 miles of wire.
    • Uses a synchronization mechanism: a 50-foot shaft driven by a 5-horsepower motor.
    • Runs early large-scale simulations for the Manhattan Project.
  • The “brains” were relays (electrically controlled mechanical switches):
    • A control wire drives a coil via current → creates an electromagnetic field → moves a metal arm to open/close a circuit.
    • Analogy: like a water faucet handle controlling water flow.
  • Limitations of relays
    • Switching speed is limited by moving parts (mass):
      • 1940s relay: up to ~50 switches/second.
      • Harvard Mark I performance: ~3 additions/subtractions/sec; ~6 sec per multiplication; ~15 sec per division; complex trig functions could take over a minute.
    • Mechanical wear and failure
      • Moving parts degrade over time (break/stick/slow down).
      • More relays → higher chance of failures.
      • Harvard Mark I: ~3,500 relays; even with a 10-year lifespan, replacing ~1 faulty relay per day during multi-day calculations was required.
    • Environmental issues
      • Machines attract insects (e.g., moth causing a malfunction), contributing to the “bug” concept.

• Vacuum tubes enabled electronic computing and major speed gains

  • Driving need: faster, more reliable switching than relays.
  • Key developments in vacuum-tube technology
    • 1904: John Ambrose Fleming invents the thermionic valve (first vacuum tube)
      • One electrode is heated → emits electrons (thermionic emission).
      • Current flows only under correct charge conditions (acts like one-way flow).
    • 1906: Lee de Forest adds a third electrode → the triode
      • Positive control electrode permits electron flow; negative stops it.
      • This effectively creates an electrical switch (open/close) without moving parts.
  • Advantages over relays
    • No moving parts → less wear.
    • Much faster switching: thousands of times per second.
  • Remaining issues
    • Vacuum tubes are fragile (can burn out like light bulbs).
    • Initially expensive; early computers may need hundreds/thousands of switches, though by the 1940s costs/reliability improved for government-scale projects.

• Early programmable electronic computers used vacuum tubes

  • Colossus Mk 1 (completed Dec 1943, Bletchley Park, UK)
    • Designed by Tommy Flowers.
    • Used to decrypt Nazi communications.
    • Contains 1,600 vacuum tubes; ten Colossi built.
    • Regarded as the first programmable, electronic computer.
    • “Programming” method: plugboards
      • Hundreds of wires plugged in like telephone switchboards to configure operations for a specific task.
  • Alan Turing connection
    • Mentioned that two years prior, Turing created the electromechanical Bombe at Bletchley Park to break Enigma codes.
    • The Bombe is explicitly noted as not technically a computer (per framing) and deferred for later discussion.
  • ENIAC (completed 1946, University of Pennsylvania)
    • Designed by John Mauchly and J. Presper Eckert.
    • Described as the first truly general-purpose, programmable electronic computer.
    • Performance: 5,000 ten-digit additions/subtractions per second.
    • Operational lifespan: about 10 years; estimated to perform more arithmetic than the entire human race up to that point.
    • Limitation: frequent vacuum tube failures → typically operational only about half a day at a time.

• Transistors replaced vacuum tubes, enabling smaller, cheaper, more reliable computers

  • By the 1950s, even vacuum-tube computing had limits.
  • Example context: US Air Force AN/FSQ-7 (completed 1955), part of the SAGE air defense system (referenced for later).
  • Need: a new electronic switch to reduce cost/size and improve reliability/speed.
  • 1947: Transistor invention (Bell Labs)
    • John Bardeen, Walter Brattain, William Shockley invented the transistor.
  • Core transistor idea (simplified)
    • Like a relay/vacuum tube, it’s a switch controlled by a control wire.
    • Uses semiconductor material whose conductivity changes with applied electrical charge.
    • Control is applied via a “gate” electrode; changing gate charge controls current flow (again, compared to a faucet analogy).
  • Why transistors mattered
    • Early transistor could switch ~10,000 times per second.
    • Solid-state (no fragile glass/moving parts) → improved reliability.
    • Much smaller components → smaller, cheaper computers.
  • Example milestone
    • IBM 608 (1957): first fully transistor-powered commercially available computer
      • Contains ~3,000 transistors.
      • Performance: ~4,500 additions/sec, roughly 80 multiplications/divisions/sec.
    • IBM transitioned products to transistors, bringing computing into offices and later homes.
  • Modern scale mentioned
    • Transistors today can be under 50 nanometers (paper thickness ~100,000 nm).
    • Switch millions of times per second; can run for decades.
  • Silicon Valley connection
    • Semiconductor development occurred in Santa Clara Valley, leading to Silicon Valley.
    • William Shockley moved there; his company contributed personnel lineage to Fairchild Semiconductors, whose employees later founded Intel.

• Bridge to next topic

  • The video ends by stating the next challenge: how to do computing using transistors—without motors/gears—and indicates this will be covered in upcoming episodes.

---

Methodologies / instruction-like elements (as presented)

Relay operation (control-to-switch mechanism)

1. Apply current to a control coil (via control wire).
2. The coil creates an electromagnetic field.
3. Electromagnetic attraction moves a mechanical arm.
4. The arm closes/opens a circuit, routing electricity to controlled components (e.g., motors/counters).

Vacuum tube / triode switching (control electrode gating current)

1. Heat the cathode electrode to emit electrons (thermionic emission).
2. Apply charge to the control electrode:
   • If control electrode is positive → allows electrons to flow to create current.
   • If control electrode is negative → prevents electron flow (no current).
3. Result: an electronic on/off switch with no moving parts.

Colossus “programming” via plugboards

• Configure the computer by plugging hundreds of wires into plugboards.
• This wiring sets which operations the machine will perform for the chosen computation.

---

Speakers / sources featured (named)

• Narrator / instructor voice (Crash Course style; not individually named)
• Thought Bubble (segment narrator/illustrated explainer; not individually named)
• Tommy Flowers (designer of Colossus Mk 1)
• John Ambrose Fleming (inventor of the thermionic valve/vacuum tube, 1904)
• Lee de Forest (inventor of the triode via third electrode, 1906)
• Grace Hopper (quoted about “bugs in it”)
• Alan Turing (mentioned; created the Bombe at Bletchley Park)
• John Mauchly (ENIAC designer)
• J. Presper Eckert (ENIAC designer)
• John Bardeen (transistor inventor, Bell Labs)
• Walter Brattain (transistor inventor, Bell Labs)
• William Shockley (transistor inventor; later Silicon Valley figure)
• Bell Laboratories (Bell Labs) (institution source for transistor invention)
• IBM (institution; IBM 608 and broader transition to transistors)
• US Air Force (institution context for AN/FSQ-7 and SAGE)