ild a kit computer (Shurkin 307; Campbell-Kelley and Aspray; Triumph). The first thing Roberts had to do was figure out how he was going to build a computer. His first decision was on what processor to use. Intel had recently released the 8080 microprocessor, which was similar to the 8008, but infinitely more powerful. This is the processor that Roberts decided to use in his machine (Freiberger and Swaine 31; Kidwell and Ceruzzi 92-93). There were other candidates, though. Motorola had recently released the 6800 processor. Roberts turned that down because one of his friends told him that he had tried to implement BASIC, a high level programming language, on the processor and it was nearly impossible (Freiberger and Swaine 31). The largest factor in Roberts decision, though, was the price factor. Intels processors sold for $360 a piece, but Roberts was sure he could talk them down. He didto $75 a piece. This was an amazing deal, but of course it required Roberts to buy in volume (Freiberger and Swaine 31; Shurkin 308; Triumph). This was no problem, though, because Roberts needed to sell a lot of something to stay in business. Roberts immediately began work on his design. He worked out a rough sketch of what he wanted and then handed the designs over to his engineers. Roberts wanted was an open architecture, he wanted his computer to be easily expandable. What this required was a number of uniform expansion slots and a way to connect them all together. The solution to this was a bus structure. A bus is a channel for data or instructions on a computer. Normally it is a parallel channel with many different signals passing simultaneously. The bus on Roberts computer had 100 separate channels and it was later dubbed the s-100 bus (Freiberger and Swaine 32-33). While the engineers were finishing up the physical designs of the computer it was still without a name. Many suggestions were made, but Roberts refused them all. What finally happened i...
