1. A Brief History of the NES
In broad terms, the NES (“Nintendo Entertainment System”) is a home
video game console designed to connect to a CRT television. It uses
interchangeable, read-only cartridges (“Game Paks”) to store games,
since the system has no permanent means of storage. Input comes in
the form of two controller ports on the front of the system; the
standard NES controller has a 4-way, cross-shaped directional pad
and four additional buttons, labelled Start, Select, A, and B.
An NES console with one controller.
Photo by Evan Amos.
The NES was first released in the U.S. in 1985; no one is sure exactly when, but most sources estimate it was in October of that year. Nintendo continued to sell the NES in the U.S. until 1995, well past the launch of its successor, the Super NES, in 1991. During that ten-year period, nearly 700 officially-licensed games were released for the NES, produced both by Nintendo directly and by a wide range of third-party developers like Konami and Capcom. It was the best-selling video game console of its generation, David Sheff. Game over: how Nintendo zapped an American industry, captured your dollars, and enslaved your children. 1st edition. Random House New York, 1993. ISBN: 0679404694. with “playing Nintendo” becoming synonymous with video games in the U.S. in the same way that “playing Atari” had been in the previous decade.
The NES did not spring fully-formed from a Nintendo R&D lab. A unique series of failures and innovations led to the system becoming a powerhouse in the home video game space. So, before we start to work with the NES, it is important to take a step back and discuss where the system came from. The historical background that led to the NES had important implications for the system’s design and capabilities, which will ultimately impact the games that can (and can't) be made for the system.
The American Experiment
Nintendo began in 1889 in Kyoto, Japan, as a hanafuda playing cards manufacturer. Sept. 23, 1889: Success is in the Cards for Nintendo. When Hiroshi Yamauchi, the great-grandson of company founder Fusajiro Yamauchi, took over the business in 1949, he began to expand the company’s product line, diversifying into new areas like taxi services, love hotels, and toys like the Ultra Hand. As Nintendo Turns 125, 6 Things You May Not Know About This Gaming Giant. Nintendo would not find breakout success until it began creating electronic toys. Nintendo began producing the Color TV-Game series of home “Pong”-clone video game consoles in 1977, and in 1980, they released the “Game & Watch” series of handheld video games. High Score!: The Illustrated History of Electronic Games.
Nintendo had released a few arcade games by this point, the most
successful being 1979’s Radar Scope
Radar Scope. Top: A promotional flyer advertising the game. Bottom: The game in action. , a space shooter game that was the #2 arcade game in Japan behind Namco’s Pac-Man. The Ultimate History of Video Games: From Pong to Pokemon - the Story Behind the Craze That Touched Our Lives and Changed the World. By 1978, Namco had already found great success in American arcades, and Taito had licensed its wildly-popular Space Invaders to Midway for the U.S. market as well. Nintendo knew that the American market presented a huge opportunity, but it would need to compete with both its Japanese rivals as well as American arcade companies like Atari.
Nintendo president Hiroshi Yamauchi thought Radar Scope would be an excellent way to start a Nintendo U.S. branch. In 1980, he gave his son-in-law, Minoru Arakawa, instructions to rent office and warehouse space for a new “Nintendo of America”, and shipped him 3,000 Radar Scope machines from Japan.
Unfortunately, waiting nearly a year to begin shipment of Radar Scope machines—plus the extra two weeks it took for them to get from the west coast to Arakawa’s New Jersey warehouse—meant that Radar Scope looked and felt dated, and American buyers were not particularly interested. Arakawa’s sales team were only able to sell about 1,000 Radar Scope machines, with the rest sitting in the warehouse gathering dust.
This was a serious problem for Nintendo. In the early 1980’s, 2,000 arcade machines represented a substantial amount of money. For the sake of comparison, during this time period Atari would provide interested customers with a “Project Materials Cost Estimate” detailing the parts required for an arcade cabinet and their costs. A typical Atari arcade game like Missile Command contained $871 in parts, and would sell to an arcade or bar for $1,995. For much more detail on the arcade business of the early '80's, see Tales from the Arcade Factory Floors. In 2018 dollars, that would be equivalent to about $2,695 in parts for a selling price of $6,170. 2,000 unsold arcade machines therefore represented over five million dollars (2018 equivalent) in sunk costs.
Hiroshi Yamauchi had an idea: what if Nintendo designed a new game
that could re-use the majority of Radar Scope’s components,
something more appealing to American buyers? Yamauchi set aside
$100,000 (1980 dollars) for the project, and handed
responsibility for coming up with the new game to a young designer
who had previously worked on the exterior shell of the Color
TV-Game series. That designer was Shigeru Miyamoto,
Shigeru Miyamoto, at E3 2013.
Photo by Jan Graber, licensed CC-BY-SA 3.0 de. who would later become famous as the creator of Mario, Link, and many more of Nintendo’s most famous game characters.
Miyamoto created a game that featured a fearless carpenter whose girlfriend had
been captured by a large ape. His inspirations included the comic strip Popeye
and the classic film King Kong. The result was Donkey Kong.
Donkey Kong. To make the game more “American”, Nintendo of America named the hero “Mario” (after warehouse landlord Mario Segale), and named the hero's girlfriend “Pauline”. Donkey Kong was enormously popular, both in the U.S. and Japan, and it set up Nintendo as an arcade powerhouse.
The Home Market
The success of the Radar-Scope-to-Donkey-Kong conversion project also led to another great idea: a home system which, like a Radar Scope cabinet, could run different game software on the same hardware. Home video game systems up to this point had been single-use machines, often built to play just one game (potentially with variants on a theme). Nintendo's own Color TV-Game could play six or fifteen varieties of Pong, depending on which model the user purchased, but nothing else. The most versatile system of the era was the Magnavox Odyssey, which could play over a dozen games by swapping out "game cards" included with the system. The cards did not contain any actual game data, though; they merely activated different parts of the system's internal circuitry, meaning the Odyssey's future expansion was extremely limited. The proposed Nintendo home system would operate the same way the arcade Donkey Kong machine did: game code would live on chips separate from the main processing components, with no pre-programmed game code on the system itself.
There was one major obstacle to creating such a system: cost. Video games, especially for the home, were marketed to families with children as a fun way to spend time together. If the machine were too expensive, parents would not consider buying it, no matter how good its games were. No parent would be willing to spend $871 for a home system, so Nintendo R&D had to find a way to make the Donkey Kong arcade cabinet for less. Yamauchi's goal was a system powerful enough to play Donkey Kong in the home, but able to sell for a mind-boggling ¥9,800 (equivalent to about $40 USD).
To meet that price target, Nintendo R&D started with the system's processor. The
Donkey Kong arcade machine used, at its core, a Zilog Z80 processor
A Zilog Z80 processor. , developed by a group of former Intel engineers who attempted to take all of the great features of the Intel 8080 series of processors and implement them in a smaller, faster, and cheaper package. The Z80 was everywhere in the Japanese computer market, powering the "MSX" standard for home computers.
Nintendo was unable to find a manufacturing partner that would agree to its strict requirements, however. The only Japanese company desperate enough to sign a deal with Nintendo was Ricoh. Ricoh was utilizing only 10% of its chip manufacturing capability at the time, so to increase business it agreed to provide Nintendo with three million processors, purchased up-front. Nintendo was going all-in on their home console bet.
Ricoh, however, did not have a license to manufacture the Z80. Securing that license
would be costly and take a long time, so Ricoh instead offered an alternative: why
not use the MOS Technologies 6502 instead?
The MOS Technologies 6502.
Photo by Dirk Oppelt, CC-BY-SA Ricoh already had a license to manufacture the 6502, an 8-bit processor with a similar performance profile to the Z80. As an added bonus, it was relatively unknown in Japan (despite powering the Apple II, Commodore 64, Atari VCS, and many other computers popular in the West). Adopting the 6502 as the core of their new system would grant Nintendo a type of copy protection, since games written for Nintendo's system would not be easily portable to competitors' systems.
The MOS Technologies 6502 was developed by a group of former Motorola engineers who attempted to take all of the great features from the Motorola 6800 series of processors and implement them in a smaller, faster, and cheaper package. (Sound familiar?) Ricoh had a license to manufacture the 6502, but it did not have a license for the "binary-coded decimal" (BCD) functionality of the chip. To avoid running afoul of MOS Technologies, Ricoh's processor cut all electrical connections between the BCD module and the rest of the chip. When MOS Technologies was purchased by Commodore, the latter attempted to sue Nintendo for license violations but gave up when Commodore engineers discovered Ricoh's fix. You can still buy 6502-based processors today from Western Design Center (WDC), and they are commonly used in industrial applications. The 6502 became available for purchase in 1975, meaning it was already ten years old when Nintendo chose to use it in its new home system. Individual 6502 processors in 1975 sold for $25 each, and Nintendo's bulk order drove the per-processor price even lower. Ricoh paired its 6502-based processor with a custom "Picture Processing Unit" (PPU) to handle drawing graphics to the screen. The PPU allowed for sophisticated graphics, supporting a 64 color palette, a 256x240 pixel resolution display, 64 hardware sprites, and a hardware-scrolling background layer.
With the internals set, all that remained was to design the outer casing of the system and give it a name.
Nintendo chose to call their new home system the "Family Computer", shortened to "Famicom".
The Nintendo Famicom.
Photo by Evan Amos. Its case was designed to make it look like a fun toy: red and white, with gold accents. The Famicom had two controllers hard-wired to the sides of the console, with a 15-pin expansion port on the front. Game cartridges were inserted vertically into a slot on top of the console.
The Famicom launched in July 1983 to great success, selling 500,000 consoles in two months despite only having three available games — Donkey Kong, Donkey Kong Jr., and Popeye. While Nintendo R&D was not able to hit Yamauchi's ambitious ¥9,800 target, the system did sell for ¥14,800 (equivalent to $65 USD), certainly cheap enough to appeal to parents.
An American Renaissance
With its popularity in Japan, bringing the Famicom to the U.S. was a no-brainer. But the American market was having issues of its own; 1983 was the year of the "Atari Crash" in the U.S., when retailers abandoned the idea of home video games after overproduction and a lack of quality control led to massive unsold inventories of Atari 2600 cartridges. U.S. retailers were not interested in buying inventory of a new video game system.
Nintendo spent two years re-designing the Famicom for a post-video-game market. It could not look like a video game machine, so its outer casing became a grey rectangle with a prominent front flap - a move meant to make it look like a VCR. The Famicom's small, colorful game cartridges became large, grey "Game Paks", mostly containing empty space but now large (and uniform) enough to bear a passing resemblance to VHS tapes. Finally, the system needed a name change. "Nintendo Entertainment System" positioned the system as part of a home theater setup, something that would fit in alongside stereo equipment.
The re-designed system was almost ready, but Nintendo felt it still needed something unique
to make it a "must-buy" product. The "Deluxe Set" included a "Zapper" light gun, perfect
for shooting games like Duck Hunt, and R.O.B., the "Robotic Operating Buddy".
Photo by Evan Amos, CC-BY-SA 3.0. This was no mere video game system - it came with a robot! The cameras that formed R.O.B.'s "eyes" could detect coded patterns of flashing light from the TV, which would instruct R.O.B. to turn left or right or move its arms up or down. R.O.B. could be used with two games, Gyromite and Stack-Up, neither of which were particularly good. But games were secondary: R.O.B. existed solely to sell the NES to American parents who had already been burned by the Atari 2600.
There is one more piece of the Famicom's re-branding to discuss before we move on to programming the hardware. A key cause of the 1983 crash was a lack of quality control or licensing for Atari 2600 games. Video games were big business, and any company that could hire a programmer or two wanted to find a way to sell Atari games. The result was a flood of games that often barely worked. Most consumers, after being burned a few times by paying full price for terrible games, stopped buying them, and the video game market imploded.
Desperate to prevent this situation for the NES, Nintendo added a chip to the NES
motherboard called "10NES"
The 10NES chip on an NES cartridge.
Photo by (you guessed it) Evan Amos. (sometimes referred to as "CIC", for "Checking Integrated Circuit"). The 10NES chip acted as a cryptographic lock and key. The console and cartridge each had their own CIC chip; the two chips would each calculate a value based on a specific algorithm, and the cartridge would send its answer to the console for comparison. If the results matched, the system would boot as normal. If they did not match, the CIC chip in the console would reset and try again. Incidentally, this is why you would often have to "blow on the cartridge" to get an NES game to work. If the connectors on the cartridge did not make a secure connection to the console, the 10NES chip would be unable to transmit its code, and it would go into a reset loop. Blowing on the cartridge actually did nothing; what fixed the problem was removing and re-inserting the cartridge to re-seat the connectors. The design of the 10NES chip as well as the specific algorithm running on it were patented by Nintendo, meaning that only Nintendo could legally manufacture 10NES chips. As a result, only Nintendo could manufacture cartridges that would work with the NES — unlicensed cartridges would not have the chip, and the NES console would refuse to boot them. Nintendo would have final say on what games were released for its platform.