The first electronic digital computer was scrapped, then rebuilt for 350,000 dollars
In 1939 a physics teacher at Iowa State and his graduate student set out to build an electronic digital computer. It took them five years. In the 1940s the machine was taken apart and thrown away. Decades later, a team spent four years and 350,000 dollars recreating it as a tribute to its inventors.
That machine, the Atanasoff–Berry computer, is often treated as the starting point of computer engineering, the discipline that blends electrical engineering, electronics and computer science to design both hardware and the software that runs on it. John Vincent Atanasoff and Clifford Berry drew on physics and mathematics as much as circuitry, a mix the field still demands.
The personal computer needed a chain of semiconductor breakthroughs first. At Bell Labs in 1947, Brattain, Bardeen and Shockley produced the first working transistors. Carl Frosch and Lincoln Derick learned to protect silicon surfaces with oxide, Robert Noyce put a whole circuit on one chip at Fairchild in 1959, and Bell Labs demonstrated the MOS transistor in 1960. In 1971 an Intel team including Federico Faggin and Ted Hoff packed an entire processor onto a single chip, the 4004.
That same year Case Western Reserve University in Cleveland launched the first American degree in the subject; by 2015 there were 250 accredited programmes. The breadth is daunting, from processor design written in hardware description languages to wireless networks, cryptography, computer vision and embedded systems in devices as small as an FM radio or as large as the space shuttle. Many students pick a specialism in their final two years because no undergraduate course can cover it all.
Job forecasts tell a curious story. American government projections for hardware engineering growth slid from 9 per cent over 2010 to 2020 down to just 2 per cent for 2019 to 2029, before bouncing back to 9 per cent for 2025 to 2035, mostly to replace people retiring or leaving. Software roles are forecast to grow faster, while plain programming jobs are projected to shrink.
Source: Computer engineering
