RETROSPECT In the Beginning Was the CRT _{By George Dyson}

Once upon a time, there was no distinction between memory and display.

Introduced in 1897, the cathode-ray

tube brought us oscilloscopes, television, radar, computer terminals, the electron microscope, and, 110 years later, You Tube. But the hum of flyback transformers, by which so much code was written, is at an end. As the last generation of warmblooded monitors vacates our desks, let us remember that the cathode-ray tube’s contribution to digital computing began as internal memory, not external display.

Conventional CRTs display the state of a temporary memory buffer whose contents are produced by the central processing unit (CPU). Once upon a time, however, cathode-ray tubes were the core memory, and they stored the instructions that drove the operations of the CPU. This was one of those sudden adaptations of pre-existing features for unintended purposes by which evolution leaps ahead.

By 1953 there were 53 kilobytes of random-access memory in the entire world, with 5kB in the original IAS machine.

In 1945, when John von Neumann began to orchestrate the electronic computer project at the Institute for Advanced Study in Princeton, N.J. (see MAKE, Volume 06, page 190), there was no high-speed random-access memory available off the shelf. Vladimir Zworykin and Jan Rajchman at RCA agreed to supply a plug-and-play digital memory tube, christened the Selectron, an electronically switched array of 4,096 separate targets storing one binary digit each. After two years, there were still no Selectrons in existence. “They were doing things inside that vacuum that hadn’t been done before,” says Willis Ware, one of the original engineers. A 256-bit Selectron was eventually produced in limited quantities, but too late to compete with magnetic-core memory, and it barely achieved the historical footnote it deserves as a missing link between the vacuum tube and the integrated circuit.

The IAS team decided to improvise a random-access memory from commercially available parts. Existing high-speed storage was based on acoustic

178 Make: Volume 10

delay lines, developed for identifying moving targets by distinguishing radar signals that had shifted from one moment to the next. A series of electrical pulses, about a microsecond apart, were converted to a train of sound waves circulating in a long tube of mercury equipped with crystal transducers at both ends. About 1,000 binary digits could be stored in the millisecond it took to travel the length of a five-foot “tank.” The delay line spawned the first generation of serial-access stored-program electronic computers (see MAKE, Volume 08, page 178), although “its programming,” as Alan Turing’s supervisor Max Newman noted, “was like catching mice just as they were entering a hole in the wall.”

If you wanted one particular bit, you had to wait a full millisecond and catch it as it went by. How could you read or write any bit at any time? Researchers at MIT’s Radiation Laboratory had noted that digital information could be stored as charged spots on the face of ordinary cathode-ray tubes, as long as the pattern was regenerated a few times a second by a trace from an electron beam. The spots become positively charged (i.e., deficient in electrons) as a result of secondary electron emission by the phosphor, and the state of an individual spot could be distinguished by briefly “interrogating” that location and noting the character of a faint secondary current, of less than a millivolt, induced in a wire screen positioned close to the outside face of the tube. “Thus the phosphor containing the various charge distributions is capacitively coupled to the wire screen,” the IAS team reported, “and it is then possible by focusing the beam at a given point to produce a signal on the wire screen.”

Frederick C. Williams, after working on pulse-coded IFF (Identification Friend or Foe) radar systems in England and the United States, had developed a serial-access cathode-ray memory tube in 1946. In June 1948, he constructed a small computer at Manchester University, under the direction of Max Newman and assisted by Alan Turing, that demonstrated CRT-based storage and a rudimentary stored program. The so-called “Williams tube” memory was highly sensitive to electromagnetic disturbances, and was plagued by the presence of an electric traction line that produced stray magnetic fields.