IBM was in the mechanical data processing business long before the emergence of electronic computers. For decades the company had produced machines that counted, tabulated and measured with gears, cogs and relays. In 1946, it began an important transition away from mechanical computing with the debut of its IBM 603 Electronic Multiplier.

The 603 was the world’s first mass-produced electronic calculator. It employed 300 vacuum tubes in its calculating unit, which was connected to an IBM Type 520 punch-read unit that fed numbers into the calculator and returned results via paper card. The calculating unit could multiply two six-digit numbers 10 times faster than other calculating machines available at the time, performing as many as 6,000 calculations per hour. As the IBM 603 manual explained: “The Electronic Multiplier makes use of recently developed electronic circuits” to transform the “burdensome and usually slow-speed process of computing products to an automatic high-speed process.”

Commercial interest in the 603 came as something of a surprise. As recounted in The Maverick and His Machine: “To IBM’s astonishment, customers liked the 603 and placed orders for it. No one at IBM had predicted such success. For the first time, [IBM] got the message: Customers will buy electronic products.”

The 603 marked the first executive project, and something of a business gamble, for Thomas Watson Jr., the eldest son of IBM’s first CEO, who would succeed his father in the position. Determined to keep IBM at the forefront of technological innovation, he pushed for the machine’s creation — and its success solidified Thomas Watson Sr.’s belief in the future of electronic products.

Released in 1946, the 603 grew out of a decade of experimentation. As early as 1936, IBM engineer A. Halsey Dickinson was researching computation that used vacuum tubes. He would later invent the first method to perform electronic addition and subtraction. Vacuum tubes held great promise for researchers because they could be switched on or off quickly in a “flip-flop circuit,” enabling them to store and convey binary data. In 1942, IBM engineers Byron Phelps and C. A. Bergfors created an electronic calculator prototype in an attic at the Endicott Engineering Laboratory. It drew on key breakthroughs in vacuum tube technology to perform its instantaneous calculations.

But the 603 represented far more than a technical achievement. It unveiled the speed and power of electronic computation, pointed the way to the future of IBM’s business, and laid the groundwork for an entire industry. It wouldn’t be long before the era of electromechanical computing would give way to fully electronic devices based on vacuum tubes, transistors and integrated circuits.

No sooner was the 603 released than electrical engineer Ralph Palmer was tasked with a redesign. He focused his energies on creating a machine that would improve the 603 in two significant ways — the ability to divide and cross-punch. The result was the greatly enhanced 604. (Palmer would later contribute to the design of the IBM 701, be named an IBM Fellow in 1963, and receive the Computer Pioneer Award from the IEEE Computer Society in 1989.)

The 604 boasted a number of technological innovations. Its construction was modular, featuring “pluggable units.” This meant that each tube could be tested prior to its installation, speeding the manufacturing process and simplifying servicing of the machines. Malfunctioning machines could be quickly diagnosed and repaired.

The 604 was also the first machine to use miniature tubes, which were stacked vertically at greater density, giving the machine more power in a smaller footprint. To have greater control over tube innovation and development, Palmer founded a vacuum tube research facility on the bank of the Hudson River. The property — affectionately known as the “pickle factory” because it had been purchased from vegetable and jam merchants — is where IBM began establishing practices for cleanliness and quality control in the production of digital devices; these standards would be adopted across the industry. 

The net result was a machine as much as seven times faster than the 603’s predecessor, the 602A. One of IBM’s earliest computer scientists, Herb Grosch, marveled at the machine’s utility. “You could put together counters of various sizes, perform a completely flexible string of commands pipeline style, carry data over from card to card — and you could divide!” he said. And while the complexity of one of the most common problems at the time, matrix arithmetic — solving sets of simultaneous linear equations — would often bog down previous machines, “The 604 ate the problem for breakfast.” 

The ad campaign behind the launch of the 604 focused on the interests of the general public, touting the transformative power of electronic computing and positioning the machine as a tool that would break speed barriers and help mankind reach new heights.

The 603 and 604 were quickly recognized as the standard of computing excellence for business and inspired a raft of innovation among IBM collaborators. In 1948, engineers at Northrop Aircraft were looking for ways to accelerate the sophisticated calculations that took weeks to complete by hand. They borrowed a 603 and figured out how to connect it to an IBM Type 405 accounting machine. IBM drew on the Northrop prototype to develop the IBM Card-Programmed Electronic Calculator (CPC) in 1949. By 1955, 700 CPCs capable of several thousand operations per second had been sold, largely to aerospace firms.

IBM produced more than 5,700 units of the 603 and 604 between 1946 and 1958, and the machines became fixtures in any environment that required advanced computation. Their release meant that, for the first time, industry had the same access to technological power as researchers or government agencies. And smaller research labs without access to Harvard’s Mark I could take on problems that were previously impossible.

The ubiquity of the machines even led to appearances in popular culture. Two scientists in the comic series The Adventures of Tintin are seen working with the 604 to send the main character to the moon. More important, it set IBM on a course to propagate computing in all areas of business and society.