When the NES was conceived, the 6502 CPU wasn’t well-known in Japan — the Z80 was much more popular. So why does the NES use a 6502? (thread)
Masayuki Uemura was tasked with developing a programmable game console. His team, Nintendo R&D2, had experience creating custom chips for Block Kuzushi, a Breakout-style TV game.
Meanwhile, Gunpei Yokoi (Nintendo R&D1, inventor of Game Boy) was churning out Game & Watch units by the truckload. Uemura’s team was left understaffed due to defections.
Uemura wanted to design custom chips for the programmable console. But Japanese chip manufacturers were already at capacity due to the personal computing boom.
Only one manufacturer had capacity: Ricoh. They were known as a copier/fax company, but they could manufacture at volume.
Ricoh’s chief engineer had worked with Uemura on Nintendo’s TV games, and convinced him to to use the 6502 for the new game console. Ricoh already had a license to manufacture the 6502.
The 6502 had another benefit to Nintendo: It would make the console harder for competitors to reverse-engineer, since the CPU was relatively unknown in Japan.
In the NES, the 6502 core is inside the Ricoh 2A03 chip, which also contains the APU (Audio Processing Unit). The 6502’s decimal mode was removed to free up room on the chip.
There was one notable NES developer with prior 6502 experience: Satoru Iwata, who learned machine language on his Commodore PET while going to school. He approached Nintendo in 1983 to make games for their new console.
Nintendo was close to inking a deal with Atari to distribute the NES, and in return Nintendo would port Atari games. Iwata delivered Joust in two months, but the deal didn’t go through. Nevertheless, he became the de facto 6502 expert for Nintendo while working at HAL Laboratory.
In 2002, Satoru Iwata became CEO of Nintendo — which shows how far a career in 6502 programming can take you!