Gwen Bell’s Introduction:
Gwen Bell introduces Bob Noyce, Vice Chairman of Intel, as the “kind of, if not father, sort of wonderful uncle of the whole industry in Silicon Valley.” She highlights the 25th anniversary of the integrated circuit invention and recommends Tom Wolfe’s article in Esquire for more insights about Noyce.

Robert Noyce’s Speech:

Noyce’s Gratitude and Humor:
Noyce expresses his pleasure and honor to be present at the special occasion and thanks the audience for their generous donations, hoping some funds remain for the museum.

Publicity and Tom Wolfe’s Article:
Noyce acknowledges the recent publicity surrounding the microchip, including Tom Wolfe’s article in Esquire. He clarifies that Wolfe is primarily a novelist and not primarily a historian, emphasizing that not everything in print is necessarily true.

Comparisons to the Aircraft Industry:
Noyce mentions an article in the IEEE proceedings comparing the progress in the computer and aircraft industries. He uses a simile by Bill Davidoff, Intel’s marketing manager, comparing the progress of integrated circuits to the hypothetical advancements in the automobile industry.

Impact of Integrated Circuits and Computers:
Noyce emphasizes the profound impact of integrated circuits and computers on various aspects of society. He draws a parallel between the current advancements in the computer field and the advanced education system in medieval times, highlighting the importance of communication skills, quantitative skills, and order in the universe.

Retrospective Thoughts:
Noyce reflects on the relatively young age of the semiconductor business compared to established companies like Chrysler, which is 60 years old.

The Semiconductor Market in the 1950s:
In 1954, the semiconductor market was relatively small, with sales of only $25 million. DOD research contracts accounted for more funding than product sales. The market grew rapidly, reaching $550 million by 1960. Diodes and transistors each accounted for about half of the market. Germanium was the dominant semiconductor material, with silicon taking a smaller share.

Early Transistor Development:
During the 1950s, there was a significant effort to develop new and improved transistors. Germanium alloy transistors were commonly used, involving melting indium onto germanium to create a P-type doping. Researchers also studied the properties of germanium and discovered that heating and cooling it could change its conductivity from N-type to P-type. The realization that indium acted as a getter to remove impurities from the germanium surface prevented this conversion in transistors.

Diffusion and Photoengraving Techniques:
Diffusion emerged as a method for introducing impurities into semiconductors, providing control over the depth dimension. Photoengraving was adopted from color television technology to create precise patterns in semiconductors. This technique replaced earlier methods of making diffused transistors, such as using masks and slots.

The Invention of the Planar Transistor:
Transistors were unreliable due to impurities on their surfaces. Jean Herny’s planar transistor addressed this issue by leaving silicon dioxide on top of the transistor during diffusion. This oxide layer acted as an insulator, protecting the transistor from contamination.

Environmental Influences on Transistor Development:
The Air Force’s Project Tinker Toy aimed to miniaturize electronic components for use in missiles. The concept of molecular engineering sought to create electronic functions by combining specific molecules. The need to reduce manual labor in transistor assembly drove innovation.

The Genesis of the Integrated Circuit:
Printed circuit boards offered an alternative to manual wiring of transistors. Photoengraving ensured precise placement and reproducibility of circuit elements. The planar transistor enabled the addition of conductive layers without short-circuiting.

Isolation Techniques for Integrated Circuits:
Junction isolation used back-biased diodes to electrically isolate circuit elements. Etching and gluing techniques were also explored for isolation. Intrinsic isolation aimed to convert silicon between elements into an insulator.

Challenges and Innovations in Isolation Techniques:
Intrinsic isolation faced practical challenges and did not work effectively. Kurt Lehovac and Jay Last had previously conceived of junction isolation and etching techniques. Robert Noyce received a patent for intrinsic isolation, but it had limitations.

Challenges in Transistor Manufacturing:
Leakage current and short lifetimes of early transistors limited their practical use. Junction isolation became the preferred method for transistor fabrication. Other circuit elements like resistors and capacitors were easily made from semiconductors, but inductors remained a challenge.

Early Integrated Circuits:
Initial integrated circuits were slow and used simple logic forms. The low yield of individual transistors made it difficult to combine them into complex circuits. Market opposition arose due to the fear of losing proprietary design positions. Worst-case design practices added further challenges.

Moore’s Curve and Cost Considerations:
Moore’s law predicted a doubling of circuit complexity every year. Assembly and test costs were significant, with a minimum cost point determined by the balance between yield and complexity. Larger wafers and die sizes helped reduce defect density and increase yield. Miniaturization allowed for more circuits on a given area and reduced power consumption.

Methods for Increasing Circuit Complexity:
Increasing the size of wafers and die areas reduced the density of defects. Smaller feature sizes allowed for more circuits on a given area and lower power consumption. Advances in photolithography and other manufacturing techniques enabled the production of complex integrated circuits.

Speed and Dimensions:
Smaller circuits lead to faster speeds, making them more capable. Current production circuits are at 2 microns, approaching 1 micron, following a trend predicted in 1978. Dimensions are comparable to biological structures like neurons, opening possibilities for brain-like functions.

MOS vs Bipolar and Epitaxy:
MOS and bipolar technologies emerged, expanding circuit possibilities. Epitaxy allowed for impure underlying material with pure semiconductor on top, enhancing circuit performance.

Complexity and Challenges:
Early integrated circuits were simple Boolean functions and gates. As complexity grew, concerns arose about managing numerous leads and optimizing lead-to-gate ratios. Making more complex circuits meant fewer units produced, raising questions about the viability of large-scale manufacturing.

Memory as a Focus:
Memory was identified as a key area for VLSI development due to its high demand from computer designers. VLSI efforts concentrated on memory technologies.

Memory Cost Reduction and Complexity:
Robert Noyce explained how the cost of semiconductor memories decreased significantly over time, enabling the use of more complex and higher-capacity memory chips.

Microprocessor Development:
The high cost of custom integrated circuit designs led to the concept of microprocessors, which are programmable and can be used in various applications.

Microprocessors in Early Devices:
Initially, microprocessors were used in simple controllers and calculators, but their capabilities grew over time. Noyce mentioned an example where he offered a PDP-8 equivalent for $10, highlighting the potential cost-effectiveness of microprocessors.

Transition from NMOS to CMOS:
Noyce predicted the shift from NMOS to CMOS technology due to its ease of design and lower power dissipation.

Bipolar vs. CMOS:
As dimensions became more controllable with advancements in lithography and processing techniques, the advantage of bipolar technology over CMOS diminished.

Fundamental Limit:
Noyce discussed the fundamental limit of distinguishing between a zero and a one in computer-like equipment due to thermal noise. He suggested that operating at lower temperatures could potentially overcome this limitation.

Moore’s Law and Microcircuit Voltage:
As microcircuit dimensions decrease, the ideal voltage for operation also decreases.

Packing More Transistors:
Reducing linear dimensions by a factor of 10 allows for a 100-fold increase in the number of transistors packed in the same area.

Exponential Growth of Transistors:
Continuing the historical trend, the number of transistors on a chip could increase by another factor of 10 in the coming decade.

Potential of a Billion Transistors on a Chip:
Within a decade, it may be possible to fit between 100 million and a billion transistors on a single chip.

Challenge of Utilizing Increased Computing Power:
The challenge lies in finding innovative applications for this vast computing power.

Avoiding the Trap of Limited Imagination:
Just as early prognosticators underestimated the world’s need for computers, we should not limit our imagination regarding the potential uses of billions of transistors.

Electronics as a Marvelous Tool:
The true potential of microelectronics lies in our ability to conceive creative and groundbreaking applications for this technology.