Minuteman Missile Program
The Minuteman missile program was the US Air Force's solid-fuel intercontinental ballistic missile effort, and its guidance computers were among the earliest large applications of transistors and then of integrated circuits. Minuteman I made its first test launch in February 1961, and the improved Minuteman II, developed in the early to mid 1960s, used a redesigned guidance computer built around integrated circuits. For the semiconductor industry, the program mattered less for the missile than for what it bought: a steady stream of orders for tens of thousands of high-reliability devices, at a time when almost no one else would pay for integrated circuits.
The guidance system had to be accurate, compact and dependable at the same time. A solid-fuel missile can be kept ready for launch in a silo for years, which meant its electronics had to remain reliable during long storage and still work in the shock and vibration of launch. The guidance computer, built by Autonetics, a division of North American Aviation, took inertial sensor data and computed the trajectory. The Minuteman I computer was constructed from discrete components, using silicon transistors and diodes, with a rotating magnetic disk for storage. Every part had to be screened, because even a small chance of failure per device would produce a poor chance of success across many thousands of parts.
The shift to integrated circuits in Minuteman II was a major commercial event for the new technology. The D-37C computer for Minuteman II used integrated circuits for much of its logic, supplied in large part by Texas Instruments, while still including a count of about 2,000 discrete transistors and diodes. The precise percentages are quoted differently by different sources, so it is safest to say that Minuteman II was one of the biggest single buyers of early integrated circuits, together with the NASA Apollo program, and that the two government programs accounted for a very large share of early IC demand. The volume mattered: as Texas Instruments and other firms increased output to meet specifications, the cost of each circuit fell sharply, which opened the door to commercial customers.
The program also enforced a discipline of quality that spread through the industry. To qualify parts, the Air Force and its contractors required traceability, statistical sampling, burn-in and rejection of whole lots when failures appeared. Manufacturers learned to understand failure modes such as contamination, bond defects and leakage, and to correct them in the process. The practices of clean assembly, failure analysis and documentation that later became normal in semiconductor plants developed partly under such pressure.
Misconceptions are common about what Minuteman did. It did not invent the integrated circuit, which came from Kilby at Texas Instruments and Noyce at Fairchild before the program ordered it at scale, and it was not the only large buyer, since Apollo took a comparable role. Nor is it correct to say that the early ICs were cheap; their price was high and fell because of the volumes the programs supported. Detailed numbers, such as transistor counts and the fraction of purchases that came from one supplier, vary in popular accounts and should be treated as approximate.
The legacy is a model of how defense acquisition can underwrite a commercial technology. The programs created a dependable early market for integrated circuits, but the missiles themselves stayed in service for decades with only gradual upgrades, while the chips inside them moved on to calculators, minicomputers and eventually personal computers.
Read the Minuteman missile program as the customer that made integrated circuits worth building: its guidance computers demanded reliability that the factories had to learn to deliver, and the volume paid for the learning.