Billions of microscopic transistors integrated onto individual chips of Silicon. And one computer may have a few hundred of these chips doing different things.
Vacuum tube computers, even the largest, rarely had more than 10,000 tubes.
Chat with our AI personalities
As most plastics used in specimen tubes will melt if sterilized with heat the plastic tubes are normally single use
Pneumatic tube
NO DIFFERENCE they all use the basic three units. I/O, ARITHMETIC and PROCESSOR. The difference is enhancement of the products. Speed buss with more bits per byte I/O speed and a co processor to alleviate overhead of I/O management.
I think it draws about 4 Lbs of vacuum.
Mostly by considering what components were already available and making tradeoff analyses as to which of the available components that could do the job were best suited to their needs. For example when Howard Aiken designed the Harvard Mark I and Mark II computers he considered both vacuum tubes and electromechanical devices (e.g. relays, clutches, odometer counters), but rejected tubes as likely to be too unreliable for the large number needed. He was satisfied with a slow but reliable machine. When Eckert and Mauchly designed the ENIAC they rejected electromechanical devices outright without any tradeoffs as too slow. They insisted on a high speed machine and chose to cope with reliability issues as they arose.