Thomas Osborne (personal computer)

“I don’t think—” “Then you shouldn’t talk.”
The Mad Hatter interrupts Alice

Thomas E. Osborne
Thomas E. Osborne

The funny thing is that the first “personal computer” in the world was (in fact) a calculator. It seems the earliest documented use of the term personal computer was in the 4 October 1968, issue of Science (Science is the academic journal of the American Association for the Advancement of Science and is considered one of the world’s most prestigious scientific journals). The issue contains a Hewlett-Packard ad for its new HP 9100A, saying: “The new Hewlett-Packard 9100A personal computer, is ready, willing, and able… to relieve you of waiting to get on the big computer… Ready to relieve you of waiting to get on the big computer… Willing to perform log and trig functions, even hyperbolics and coordinate transformations at the touch of a key. Able to take on roots of a fifth-degree polynomial, Bessel functions, elliptic integrals and regression analysis.

The development of the Hewlett-Packard 9100A started in 1965 in Palo Alto when a physicist named McMillian approached HP with a small calculator he invented. Tomas Osborne, another inventor, also approached HP with his own home-built calculator (when in 1971 the machine was patented (first application was filed in 1966), namely Osborne was specified as the inventor—see the US patent Nr. 3623156). The best features of the two products were combined, and Osborne was hired as a consultant to continue the development work in HP Labs.

Osborne's calculator prototype from 1964 (© National Museum of American History)
Osborne’s calculator prototype from 1964 (© National Museum of American History)

In the early 1960s, as a graduate student in electrical engineering at the University of California at Berkeley (he earned an M.S.E.E. degree from Berkeley in 1962), Thomas E. Osborne began thinking about the design of a desktop electronic calculator suited for calculating the very large and very small numbers encountered in scientific work. In January 1964, he formed the firm Logic Design, Inc., to develop his ideas. By late 1964, he had built this prototype keyboard and display (see the nearby image), as well as a prototype logic unit (prototype overall measurements: 16.5 cm x 21.5 cm x 40 cm). In early 1965 Osborne demonstrated his calculator to several companies (like IBM and Monroe) in his apartment, where he’d built the machine, but without success, before being hired by Hewlett Packard in the spring of 1965 as a consultant with the responsibility for developing the architecture of the HP9100A.

The 9100A Desktop Computing Calculator (see the lower image) was introduced in March 1968. It was the first, totally self-contained programmable unit of its kind, which could fit on a desk (see the HP 9100A Brochure). It included a display with three registers and a magnetic card reader. An optional printer, which fit neatly on the top of the 9100A, was offered separately. The 9100A used a PC board ROM for its algorithms, including log and trig functions. Its volatile core memory used small ferrite rings through which were woven copper wires. The initial price of $4900 was rather steep, and it was later lowered to $4400. The 9100A was about the size and weight (18 kg) of a typewriter (see the Technical Description of HP 9100A) The device and the inventor have been presented in the September 1968 issue of HEWLETT-PACKARD Journal.

Programs and data were entered either through the 63-key keyboard or by means of wallet-sized magnetic cards capable of holding up to two complete read/write memory images. Data in the 9100A were represented as decimal floating-point numbers with two-digit exponents and twelve digits of mantissa precision. Results were displayed on a 5-in electrostatic CRT in three lines (numeric only), displaying the contents of three registers. Support was provided also for a full complement of arithmetic, logarithmic, exponential, trigonometric, hyperbolic, coordinate, memory, and programming functions. The speed of calculations was remarkable for the time—typical add/subtract operations completed in 2 ms, multiply required 22 ms, square-root 30 ms, and trigonometric functions 330 ms. Conditional and unconditional branching using flags and/or arithmetic comparisons were also provided, along with program halt, pause, and single-step execution.

Hewlett-Packard 9100A Desktop Computing Calculator
Hewlett-Packard 9100A Desktop Computing Calculator

The 9100A’s performance seems insignificant today with limited internal memory that stored only 196 steps. It was truly innovative for its time, however. It was organized into a 368-word by 6-bit read/write coincident current core memory for programs and data; a 64-word by 29-bit, 800 ns, threaded core ROM for control sequences; and a 512-word by 64-bit program ROM for microcode. The latter memory utilized a 16-layer printed circuit board with inductive coupling to sense lines from reference and address lines. This unusual technology achieved a density of 1000 bits per square inch using no integrated circuits!

A lot of peripheral devices can be connected to 9100A, let’s mention only:
• 9101A Extended Memory (adding 248 registers capable of storing 3472 additional program steps)
• 9104A Tape Reader (reads data into the 9100 from punched paper tape)
• 9107A Digitizer
• 9120A Printer
• 9125A X-Y Plotter
• 9150A Display Monitor (large Screen Display, with a 17-inch diagonal CRT (cathode ray tube) for use in classrooms or for displaying to any large group)
• 9160A Optical Card Reader (inputs program steps or data to the calculator using cards marked with a soft lead pencil)

Although the HP 9100A was really a desktop computer (but yet specialized for scientific and engineering purposes), the company decided to sell it as a calculator. At the time, the perception was that a computer had to be big to be accepted by the market. Calculators were also more likely to be bought than computers, she adds. Purchasing agents were authorized to buy calculators, whereas computers required top management participation, regardless of the cost. One of the company’s co-founders, Bill Hewlett, had another reason for marketing the 9100A as HP did:
“If we had called it a computer, it would have been rejected by our customer’s computer gurus because it didn’t look like an IBM. We, therefore, decided to call it a calculator and all such nonsense disappeared.”

Hewlett-Packard soon followed up with the 9100B, an enhanced version that provided several new features. In 1972 HP introduced the HP 9830A—the most powerful model of the original 9800 series calculators. It greatly blurred the lines between traditional computers and calculators due in part to its use of the BASIC programming language and options such as terminal emulation and a hard disk drive.