Masayuki Uemura arrived at Nintendo as a salesman. He worked for Sharp, selling light-sensitive solar cells, and his Kyoto office sent him to the playing-card company to see if it wanted any. Nintendo’s engineer Gunpei Yokoi did want them, and he wanted Uemura too. Ten years later Uemura led the team that built the Family Computer, the console the West knew as the NES, and after it the Super Famicom.
From Sharp to Nintendo
According to David Sheff, who interviewed him for Game Over, Uemura was born in 1943 and grew up in Kyoto, where his family had moved to escape the wartime bombing. Money was short, and as a boy he built his own toys, including radio-controlled planes made from parts he found on junk heaps. He trained as an electronics technician at an industrial college and then joined Sharp, selling the optical semiconductors used in solar cells.
The sales call to Nintendo came at the moment its president, Hiroshi Yamauchi, had told Yokoi’s engineers to look for electronic toys. Yokoi saw that a small solar cell could work as a light sensor, and hired Uemura away from Sharp. Uemura told Sheff what drew him: “Here were these very serious men thinking about the content of play. Other companies were importing ideas from America and adapting them to the Japanese market, only making them cheaper and smaller. But Nintendo was interested in original ideas.” Obituaries give 1972 as the year he joined.
Light guns and TV games
The first product was a light gun. A beam of light from the gun struck a solar cell in the target, which could set off a spring, a roar or an exploding plastic beer bottle. Sheff reports that more than a million of these Beam Gun sets sold. Uemura and a new recruit, Genyo Takeda, then worked on Laser Clay, Yamauchi’s plan to turn Japan’s disused bowling alleys into clay-pigeon shooting ranges. The first opened in Kyoto in 1973.
When Nintendo moved into home video games, it could not build the circuit boards itself. Uemura suggested working with an electronics firm, and Nintendo made its first consoles with Mitsubishi: the Color TV Game 6 of 1977, which played six versions of a tennis game, and the Color TV Game 15. Sheff reports that a million of each were sold.
The Famicom
Yamauchi’s brief for the next machine was short. It had to be something “other companies cannot copy for at least one year”, and it had to be cheap. Rival consoles in Japan sold for ¥30,000 to ¥50,000; Yamauchi wanted ¥9,800. Uemura studied the competition and decided that games needed something the office-computer designs behind them lacked: fast, smooth movement. He spent eighteen-hour days with Nintendo’s arcade engineers, trying to work out what their circuit boards really needed.
He settled on a standard 8-bit 6502 processor and a second custom chip to draw the picture. Rival machines used Texas Instruments’ TMS9918 video chip; Uemura wanted more colours, and got 52. His team tested how many moving objects the picture chip could handle, found the first answer too low, and changed the circuit. “We had to accomplish this exactly,” Uemura told Sheff. “It was the order from the president.” He took the preliminary circuit diagrams to Ricoh and asked for chips costing no more than ¥2,000 each; Ricoh agreed once Yamauchi guaranteed an order for three million.
Yamauchi also told him to leave out the extras: no keyboard, no modem, no disk drive, and cartridges rather than easily copied floppy disks. To keep costs down, the controllers were wired into the console. One of them had a microphone. According to the Videogames Hardware Handbook, the machine’s working name was “GameCom” until Uemura’s wife suggested “Family Computer”. Genyo Takeda offered Sheff a fond explanation of why it worked: Uemura “was so much an amateur that when Yamauchi told him to make this thing, he didn’t know that it could not be done.”
The Famicom went on sale in Japan in July 1983, and 500,000 sold in the first two months. Then some games began to freeze the machine. Uemura and Yokoi traced the fault to one of the chips. Yamauchi chose to recall the consoles and fix them rather than risk Nintendo’s reputation.
R&D 2 and the Super Famicom
In 1984 Yamauchi made himself supervisor of all Nintendo’s research, divided between three groups under Yokoi, Uemura and Takeda. Uemura’s R&D 2 was responsible, in Sheff’s words, for “the Nintendo hardware itself”, with about 65 people, and also made peripherals such as the Famicom’s communications adapter for Nintendo’s network. The special cartridge chips that let later Famicom games go beyond the original hardware came from Takeda’s R&D 3.
In the late 1980s Uemura’s team began the Super Famicom. It could not play Famicom games: Sheff reports that backwards compatibility would have added at least $75 to the cost of each console. The new machine had a processor built around a 65816 core, a palette of 32,768 colours, several background layers and Mode 7, which could rotate, enlarge and shrink a background. It went on sale in Japan on 21 November 1990, and Edge later recalled how Nintendo shipped the first 300,000 consoles to shops in secret, in the early hours of the morning.
After Nintendo
Uemura retired from Nintendo in 2004 and taught game development as a professor at Ritsumeikan University in Kyoto. The university’s Center for Game Studies announced his death on 6 December 2021, at the age of 78.
Sources
- David Sheff, Game Over (1993; GamePress edition, 1999), chapter 2, “In Heaven’s Hands”: his birth and childhood, Sharp and the sales call, Yokoi, the Beam Gun and Laser Clay, Takeda, the Color TV Games and Mitsubishi, Yamauchi’s brief and price target, the 6502, the TMS9918, 52 colours, the sprite experiments, Ricoh, the missing extras and the microphone, Takeda’s remark, the launch sales and the faulty chip, R&D 2 and R&D 3; chapter “The ‘Tetris’ Syndrome”: the Super Famicom and backwards compatibility.
- N64 Magazine issue 2, May 1997, Nintendo history feature: Uemura “an ex-Sharp engineer”; the Beam Gun; the Color TV Games with Mitsubishi; “It was a team led by Uemura” that created the Famicom.
- Videogames Hardware Handbook, volume 1, second edition (Imagine Publishing, 2016), Famicom chapter: the “GameCom” name, the hardwired controllers, the microphone, and Yamauchi’s decision to recall the first consoles.
- Edge issue 123, May 2003, Super Nintendo retrospective: Uemura and R&D 2 commissioned in the late 1980s, the compatibility problem, the 32,000-colour palette, the 21 November 1990 launch and the overnight deliveries.
- Nintendo, Super NES Development Manual, Book I: 32,768 colours; Mode 7 rotation, enlargement and reduction.
- The Famicom’s launch date and the Ricoh order, and the Super Famicom’s processor: see the sources for Nintendo Entertainment System and Ricoh.
- Wireframe issue 58, December 2021, p. 23: his death in early December aged 78; his Nintendo career beginning in 1972.
- Game Informer, 9 December 2021 (web, secondary): his death on 6 December, as announced by the Ritsumeikan Center for Game Studies; joining Nintendo in 1972; retiring in 2004 and teaching game development at Ritsumeikan University. https://gameinformer.com/obituary/2021/12/09/nes-snes-creator-masayuki-uemura-dies-at-78
- Game Developer, obituary (web, secondary): his 32 years at Nintendo, 1972 to 2004, and teaching at Ritsumeikan’s Center for Game Studies. https://www.gamedeveloper.com/game-platforms/masayuki-uemura-lead-architect-of-the-nes-and-snes-has-died-at-78