Remembering Atanasoff and his ABC

Remembering Atanasoff and his ABC

How do you classify an electronic digital computing device that is neither programmable nor Turing-complete (Turing-complete means a general or universal computing machine which does not perform only specialized functions)? The sad answer is you don’t. The invention remains obscure and falls by the wayside. Only to be redeemed by the fame of a later computer  which used the very same principles to create a whole new industry. This folks, is the story of the Atanasoff-Berry computer (also called the ABC computer) conceived by a mathematics and physics professor, one John Vincent Atanasoff, at the Iowa State College, way back in 1937.

Atanasoff background

John Vincent Atanasoff
Courtesy of Iowa State University Library

Atanasoff seems to have had an ordinary schooling stint, except for an interest in Mathematics, which he developed at a very young age. The article, The Introduction to Electronic Computing in the website of North Carolina State University explains,

“John Atanasoff was born in 1903 in New York State. He first became interested in mathematics when, while he was still a boy, his father purchased a Dietzgen slide rule. He became fascinated with the device, and began to analyze the mathematics behind it. He soon taught himself algebra with the help of his mother.

After graduating from high school, he attended the University of Florida, where he received a degree for electrical engineering. He continued to get a MS in mathematics at Iowa State University, and a PhD in physics from the University of Wisconsin. During his work in getting a PhD, Atanasoff, like many others, was forced to complete long, complex systems of linear algebraic equations by hand. This was a long and arduous procedure, taking several hours to complete one system.

… After receiving his PhD, he began to experiment with other, better methods including the use of mechanical and electromechanical techniques. Then in 1937, he came up with a design of a fully electronic computer.”

Initial trial and errors

Atanasoff entered into the field of computing tangentially, when he started looking for a computing system which would reduce the drudgery of having to complete long, complex systems of linear algebraic equations by hand. He tried his hand at many solutions before he had the aha! moment.

According to the article on Atanasoff in the Iowa State University, Department of Computer Science website,

“Atanasoff’s original thought was to improve upon existing calculating machines, notably the IBM tabulator. He and A.E. Brandt, an Iowa State College professor of Statistics, made modifications to the IBM tabulator so that it could solve problems in analysis of complex spectra. This work was published in the Journal of The Optical Society of America in 1936. Unbeknown to them at the time, the authors’ work was not highly regarded by IBM, whose corporate officials saw their machine being used for purposes for which it was not meant.

His next attempt was through the construction of an analog calculator called the “laplaciometer.” The Atanasoff-Hannum-Murphy Laplaciometer, a small analog caculating machine was a success in terms of accurate calculations; but Atanasoff was still not satisfied with its reliability, as its components had to be in perfect working mechanical order to guarantee accuracy.”

Eureka on the drive to Illinois

The inside news about his eureka moment came from Atanasoff himself during his deposition in the patent suite Honeywell vs. Sperry Rand. The article, The Atanasoff-Berry Computer: The First Electronic Computer in the thoughtco website describes it thus,

“His obsession with finding a solution to the computer problem built to a frenzy in the winter months of 1937. One night, frustrated after many discouraging events, he got in his car and started driving without destination. Two hundred miles later, he pulled onto a roadhouse. He had a drink of bourbon and continued thinking about the creation of the machine. No longer nervous and tense, he realized that his thoughts were coming together clearly. He began generating ideas on how to build this computer.”

Though the ideas he came up with during his drive seem obvious in hindsight, it was a significant departure from conventional thinking of the time.

“The four ideas that came together were enumerated in the article on Atanasoff in the Iowa State University, Department of Computer Science website, as follows,

1) He would use electricity and electronics as the medium for the computer

2) In spite of custom, he would use base-two numbers for his computer

3) He would use condensors for memory and would use a regenerative or “jogging” process to avoid lapses that might be caused by leakage of power

4) He would compute by direct logical action and not by enumeration (counting) as used in existing analog calculating devices.”

Building the ABC

Berry-and-ABC 1942
Berry and ABC in 1942 (Courtesy of Iowa State University)

Compared to later efforts of that era, to build computers, the ABC was a very modest effort. The article, The Atanasoff-Berry Computer: The First Electronic Computer in the thoughtco website, puts it succinctly.

“After receiving a $650 grant from Iowa State College in March 1939, Atanasoff was ready to build his computer. He hired a particularly bright electrical engineering student, Clifford E. Berry, to help him accomplish his goal. With his background in electronics and mechanical construction skills, the brilliant and inventive Berry was the ideal partner for Atanasoff. They worked at developing and improving the ABC or Atanasoff-Berry Computer, as it was later named, from 1939 until 1941.

The final product was the size of a desk, weighed 700 pounds, had over 300 vacuum tubes, and contained a mile of wire. It could calculate about one operation every 15 seconds. Today, computers can calculate 150 billion operations in 15 seconds. Too large to go anywhere, the computer remained in the basement of the physics department.”

Visit by Mauchly

The visit by Mauchly will have far reaching implications for the computer industry as a whole and to Atanasoff and Mauchly in particular.

According to the article on Atanasoff in the Iowa State University, Department of Computer Science website,

“In December 1940 Atanasoff attended the annual meeting of the American Association for the Advancement of Science (AAAS) in Philadelphia. John Mauchly was a scientist giving a presentation on his harmonic analyzer, a machine for carrying out calculations related to weather patterns. Following his presentation, JVA introduced himself to Mauchly and the two began a discussion about computing machines.

Mauchly expressed interest in the computer JVA was working on and had many questions about its operation. Atanasoff apologetically explained to Mauchly that he had been cautioned by Iowa State College and the patent lawyer Richard Trexler not to disclose information that may jeopardize the patent. But he did invite Mauchly to come to Ames and see the computer and how it functioned.

Mauchly accepted this invitation and traveled to Ames from Philadelphia, arriving on June 13, 1941. He spent four days in the company of either JVA or Clifford Berry. Mauchly saw the computer, and spent much time with JVA and Berry discussing how it was designed and how it worked. Mauchly also read parts of the 35-page documentation JVA and Berry had written on the machine and Mauchly took notes on its contents. When he asked Atanasoff for a copy, JVA declined, citing the need to protect the ideas and designs within until the patent application was finalized.

…… John Mauchly went on to build the ENIAC computer, and with J. Presper Eckert, patent the machine as the first digital electronic computer.”

World War II

While WW II, accelerated the process of building the computer in the case of the Colossus and ENIAC, it had the exact opposite effect for the ABC.

The article, The Atanasoff-Berry Computer: The First Electronic Computer in the thoughtco website claims,

“World War II started in December 1941 and work on the computer came to a halt. Although Iowa State College had hired a Chicago patent lawyer, Richard R. Trexler, the patenting of the ABC was never completed. The war effort prevented John Atanasoff from finishing the patent process and from doing any further work on the computer.

Atanasoff left Iowa State on leave for a defense-related position at the Naval Ordnance Laboratory in Washington, D.C. Clifford Berry accepted a defense-related job in California. On one of his return visits to Iowa State in 1948, Atanasoff was surprised and disappointed to learn that the ABC had been removed from the Physics Building and dismantled. Neither he nor Clifford Berry had been notified that the computer was going to be destroyed. Only a few parts of the computer were saved.”

ENIAC Computer

ENIAC, though by all accounts a more considerable effort, was a computer that worked as compared to the somewhat erratic ABC, nevertheless owed many of its path breaking innovations attributed to Eckert and Mauchly, owed their provenance to Atanasoff.

According to the snippet, ENIAC: The First Electronic Computer. Until it Wasn’t, in the,

“Being the first electronic computer involved more than bragging rights. It involved money.

ENIAC’s inventors filed for patents in 1947. They were finally issued in 1964, and patent-holder Sperry Rand sought royalties from competitors.

Honeywell and CDC objected, citing prior work by John Atanasoff, who conceived an electronic computer in 1937 and built it in 1939-1942. Significantly, ENIAC inventor John Mauchly had visited Atanasoff in 1941.

In 1967 the dispute landed in court. The ruling, recognizing Atanasoff’s earlier work, revoked Sperry’s patents. So, one of history’s key inventions is owned by…nobody.”

Rebuilding ABC

All these early computers have something in common. The penchant of later engineers wanting to re-build the machines for posterity. Like the Colossus, ABC was also re-built.

You can find the details in the article, ABC Reconstruction, 1994-1997,

“The project was initiated and championed by Delwyn Bluhm, then Manager of Engineering Services at Ames Laboratory, and George Strawn, former director of the Iowa State University Computation Center and Chair of the ISU Department of Computer Science. Bluhm and Strawn obtained initial funding from Charles Durham, a former student of John V. Atanasoff and distinguished alumni of ISU. The original ABC was built with a grant from the Iowa State College for around $5,000. The ABC Reconstruction was completed in 1997, and cost around $300,000. Funding for the project was obtained entirely through private sources.”

In spite of getting the raw end of the deal in getting the credit for invention of the computer, John Vincent Atanasoff was gracious enough to tell reporters, “I have always taken the position that there is enough credit for everyone in the invention and development of the electronic computer.”

Let’s next look at the story of ENIAC which grabbed the limelight for the better part of 25 years, as the first ‘digital electronic computer’ and then fell from grace. But, that’s next week.

Leave a Reply