Who Invented the Computer? Vannevar Bush

Who Invented the Computer? This is the fifth installment in our ongoing series.

Note: In our previous installment, we learned about the application of Lord Kelvin's (William Thomson's) various mechanical inventions, including their incorporation in the prediction of tides and the more effective use of ship-mounted large guns.

While other scientists would tinker with and apply Thomson's integrator with various results, it wasn't until 1931 that a widely practical general-purpose analyzer arrived on the scene, designed by a professor at the Massachusetts Institute of Technology, Vannevar Bush.

Like Thomson, Bush was a wunderkind — he completed his Ph.D. in electrical engineering in just eight months and went on to have arguably the largest impact on science in the United States across the entire 20th century. His accomplishments are amazing.

Over an achievement-packed career, Bush refined the "S-tube," making radio a simple and inexpensive plug-in device; and served as President Franklin D. Roosevelt's top science advisor during World War II, managing all defense contracts and initiating and managing the Manhattan Project to develop the atomic bomb.

After the war, Bush founded the defense contractor Raytheon and played a major role in establishing the National Science Foundation. He is also credited with anticipating modern search engines by 50 years in his "visionary and influential" article, As We May Think. (It's worth a read.)

The idea for a differential analyzer came about in 1928, when Bush was working with electrical circuitry and grew frustrated having to work out "tough equations." Over the next three years he and one of his students, Locke Hazen, built an analyzer that could model power networks and solve arbitrary differential equations of the sixth-order.

Realizing the power of what he and Hazen had created, Bush began to use the machine not only to analyze power transmission networks, but also to solve complex problems in physics, ballistics, and seismology.

War machine

Before the United States entered World War II, the government tasked Bush with building an extremely sophisticated analyzer, the Rockefeller Differential Analyzer Number 2 (named for the Rockefeller Foundation, which paid for its development). So important was the project that government officials deliberately spread misinformation to the effect that it had had failed and been discontinued.

During World War II, U.S. and British forced found plenty of uses for the supercharged Rockefeller analyzer, running it around the clock to calculate firing tables, map antenna profiles, and even back-trace the trajectory paths of V2 rockets — which led to the Allies discovering and bombing the Peenemunde launching site.

The Rockefeller analyzer even solved the challenge of bombing hydro-electric dams in Nazi Germany by determining the precise procedure and timing for an airplane to drop a bomb and skip it across water to sink to the base of the dam before exploding.

As with any new and useful inventions, once made public, Bush's mega-analyzer was soon copied and improved upon by other nations and organizations. One notable version was constructed by Douglas Hartree of the University of Manchester, England.

On a visit to MIT in 1932, Hartree observed Bush's analyzer in operation. Keenly impressed, he returned home and, with help from research student Arthur Porter, built his own analyzer entirely out of Meccano, the U.K.'s answer to America's trademarked Erector set. Hartree's machine wasn't as accurate as Bush's model, but it cost a lot less — $100 as opposed to $25,000.

The heyday for differential analyzers would last into the early 1960s. Over the course of that 30-year reign, they were one of the main calculating tools for solving differential equations. Eventually, digital computers became ever more powerful and inexpensive enough for most universities and research facilities to make the switch from analyzers.

Although long-since replaced by digital computers, differential analyzers are still built today — mostly by interested individuals and programs involved with teaching advanced mathematics. They are an excellent teaching tool for modelling complex problems to students and remain a fast way to solve differential equations.