The controversial ENIAC patent trial
By the mid-1960s, pharmacy long after ENIAC’s obsolescence, discount Sperry Rand was just beginning to receive royalties from the computer patent which was not granted until 1964. The company negotiated a cross-licensing deal with its largest competitor, viagra IBM, but could not reach agreements with smaller competitors such as Honeywell. This led to Sperry Rand and Honeywell suing each other in May 1967. Sperry Rand alleged that Honeywell engaged in patent infringement. Honeywell alleged that Sperry Rand had an illegal monopoly and that the ENIAC patent was unenforceable because of prior art in Atanasoff’s device. A trial led by U.S. District Court Judge Earl Larson convened from June 1971 to March 1972 in Minneapolis, which was Honeywell’s home turf.
Honeywell’s attorneys presented the Atanasoff device and Mauchly’s correspondence with Atanasoff as evidence that ENIAC technology concepts were based on Atanasoff ideas. They also argued that the ENIAC patent was filed too late (1947) because the computer was delivered to the Army two years prior. Judge Larson agreed and ruled the patent unenforceable, although he added that Honeywell would be infringing if Sperry Rand were to win on appeal.
Historians are divided on whether Larson’s decisions were good for the American computer industry. One perspective is that all companies now were free to make and sell computers without fearing the ENIAC patent, while Sperry Rand could also compete without fearing government antitrust hearings.
But was Judge Larson correct when he decided that the ENIAC patent illegally borrowed from the Atanasoff device (ABC)? And if not, then why didn’t Sperry Rand appeal on that point?
Honeywell attorneys argued that the ABC beat ENIAC to several important concepts such as using vacuum tubes and separating the processing from the memory. They also showed how Mauchly had visited Atanasoff prior to building ENIAC. Their “smoking gun” was a letter in which Mauchly asked Atanasoff’s approval for building a similar device.
Honeywell’s arguments remain problematic to computer historians even today. The ABC contained technological innovations, and Mauchly did learn about it directly from Atanasoff before building ENIAC, but ENIAC as it actually existed was unlike the ABC in many vital aspects:
- ABC was never finished; ENIAC served for many years
- ABC used vacuum tubes for logic; ENIAC used counters
- ABC contained mechanical parts; ENIAC was all-electronic
- ABC needed human intervention; ENIAC was automatic
- ABC was not programmable; ENIAC was programmable
Therefore it’s clear that despite Atanasoff’s developments and Mauchly’s letters, the ENIAC was not derived from the ABC. (In fact, it’s well-documented that Atanasoff himself borrowed ideas from Babbage and from military devices, and that Mauchly had independent ideas and studied a wide range of calculating and computing devices before visiting Atanasoff.)
There is also documentation showing that Mauchly and Eckert only made narrow claims in their patent application, but that Remington Rand and later Sperry Rand independently expanded the claims to be overly broad. Some historians believe this influenced Larson’s decision to declare the final patent unenforceable. Larson also agreed with Honeywell’s argument that the ENIAC patent was filed too late to be valid.
If Sperry Rand had appealed, the company could have earned billions in royalties, as the patent would have been valid for 17 years from 1964 — all the way to 1981, when microcomputers were already taking over the industry.
Instead, unwilling to fund the appeal and unable to predict the future, Sperry walked away. This was based on Larson’s decision, including the decision that only IBM, not Sperry, successfully created a monopoly. In the latter sense, Sperry Rand got a break on what could have been a lethal charge to its very existence.
Despite the legal nuances, history and technology exist distinctly from courtrooms. The facts show that Atanasoff did not make a working electronic computer, Mauchly and Eckert did, and it was dissimilar to the ABC.
ENIAC Patent Trial Collection (University of Pennsylvania) – Overview of the entire case with links to many legal documents
ENIAC File Wrapper (IEEE Technology and Society) – The ENIAC patent application is examined here