The history of the calculator

Before machines: counting with your hands

For most of history, "calculating" meant counting on fingers, with pebbles, or on an abacus, a frame of beads on rods. A skilled abacus user could add and subtract remarkably fast, and in parts of Asia it was a normal shop tool well into the twentieth century.

In 1617 the Scottish mathematician John Napier published a set of numbered rods, later nicknamed Napier's bones. Lay the right rods side by side and a hard multiplication becomes a few easy additions. Within a few years the slide rule appeared, built on Napier's other big idea, logarithms. Engineers would carry slide rules for the next 350 years.

The first mechanical calculators

In 1623 Wilhelm Schickard, a professor in Tübingen, Germany, built what he called a Calculating Clock. It added and subtracted with gears, and he described it in letters to his friend Johannes Kepler. The machine was lost in a fire and the idea was forgotten. Historians only rediscovered his drawings in the twentieth century, so for a long time someone else got the credit.

That someone was Blaise Pascal. In 1642, still a teenager, he watched his father, a tax official, grind through endless sums. Pascal built a brass box of dials and gears, the Pascaline, that could add and subtract. It was fiddly and expensive, but he made a few dozen, which was enough to make it famous.

Thirty years later Gottfried Leibniz wanted a machine that could also multiply and divide. His Stepped Reckoner used a drum with teeth of different lengths. The machine itself was unreliable, but the stepped drum was such a good idea that calculator makers used it for two centuries.

Calculators go to work

None of those early machines were sold in any number. That changed in 1820, when Thomas de Colmar patented the Arithmometer, sturdy enough for insurance offices and banks to buy. In 1887 Dorr Felt's Comptometer added a full keyboard, so a trained operator could add a column of numbers as fast as she could press keys. Electric motors followed in the early 1900s, and by the 1950s offices were full of heavy, clattering desk calculators from Friden, Marchant and Monroe.

The most charming machine of the era was the Curta (1948), a hand-cranked calculator the size of a pepper grinder. Rally drivers used them to work out times and distances well into the 1970s.

The electronic revolution

In 1961 the British Bell Punch Company launched ANITA, the first all-electronic desktop calculator. It used glowing tubes instead of gears, so it was silent, fast and expensive. It had the electronic market almost to itself for two years.

Then transistors arrived. Friden's EC-130, announced in 1963, had no tubes at all, showed its numbers on a small screen, and used the stack-style "Reverse Polish" entry that Hewlett-Packard later made famous. Sharp in Japan was close behind, and within a few years manufacturers were switching to integrated circuits, putting dozens of transistors on one sliver of silicon. Each step made calculators smaller, cheaper and more reliable.

A calculator that fits in your pocket

By 1970 the obvious goal was to put a whole calculator on a single chip. A Dallas company, Mostek, did it for the Japanese firm Busicom. In 1971 Busicom used that chip in the LE-120A "Handy", one of the very first truly pocket-sized calculators, with a red LED display and ordinary batteries.

Busicom also asked a young company called Intel to design chips for a printing calculator. Intel's engineers suggested something different: one general-purpose chip that software could tell what to do. The result was the Intel 4004, the first commercial microprocessor, and it first shipped inside Busicom's 141-PF calculator in 1971. In a real sense, the chip inside every computer and phone today began as a calculator part.

The scientific calculator kills the slide rule

Hewlett-Packard already made a powerful desktop scientific calculator, the HP 9100A. Bill Hewlett wanted one that fit in his shirt pocket. In 1972 HP released the HP-35, named for its 35 keys. It did trigonometry, logarithms and powers, and cost $395, a lot of money then. Engineers bought it anyway, and the slide rule went from essential to museum piece within a few years.

Prices fell fast. Texas Instruments' TI-30 (1976) made scientific functions affordable for students. In 1981 HP launched the HP-12C financial calculator, still on sale today. In 1985 Casio's fx-7000G became the first graphing calculator, and graphing calculators became a fixture of math classrooms.

From devices to software

Today most calculating happens in software: the calculator app on your phone, spreadsheets, and websites like this one. A modern phone can do more arithmetic in a second than all the mechanical calculators ever built could do in a lifetime. The goal hasn't changed since Pascal, though: get the right answer quickly, and trust it.

That last part is why Calcist shows its working. A calculator you can check is more useful than one you just have to believe. Try the scientific calculator (the HP-35's descendant), the basic calculator with a running tape (very Comptometer), or the fraction calculator, which keeps answers exact the way no slide rule could.

Calculator history timeline

Dates are for when each machine was introduced or first described. Historians sometimes differ by a year, depending on whether they count an announcement, a prototype or the first sales.