[UPDATE: A previous version of this article misrepresented the developers of the 6502 and 65816. They were both developed by Ricoh. Goomba Stomp regrets the error.]
In recent years, Nintendo has been known more for the eccentricity of its hardware more than anything else. From the Wii’s motion controls, to the Wii U’s dual-screen gameplay, and the Switch’s portability, many analyses have focused on the physical design of Nintendo’s hardware while ignoring the systems’ most critical underpinnings: its architectural design. While Sony and Microsoft have been locked in a fierce war over console specifications for the past three generations–pitting each others’ machines in an ever-evolving battle for supremacy–Nintendo has focused on the software.
That’s a shame, though. As the progenitor of the modern games industry and a talented hardware manufacturer, Nintendo deserves a closer look at technology behind their consoles and what makes them unique. In part one of a three part series, we dive deep into Nintendo’s design choices for their first three major consoles: the NES, SNES, and Nintendo 64.
Nintendo Entertainment System (NES): The E-cycled Titan
Released as the Famicom (Family Computer) in Japan, the NES was, by no means, a technical powerhouse. Powered by the legendary, but aging, Ricoh 6502 (also known by its chip ID, the 2A03) clocked at 1.79MHz (1.66 MHz PAL), which had powered systems from the Atari 2600 to the Commodore 64, the NES utilized an aging chipset to accomplish Nintendo’s core goal with the system: creating a faithful home console port of the arcade hit, Donkey Kong. It also succeeded in attracting the talents of a young programmer, Satoru Iwata, one of Japan’s few programmers to have extensive experience with the 6502.
The core Ricoh 6502 accomplished this, and more, by creating a bevy of interesting and creative games that peaked with 1985’s Super Mario Bros. However, Nintendo wasn’t finished there. The NES’ flexible design, enhanced by a dual ROM chipset and expandability through more expensive cartridges, allowed the NES to tackle much bigger, and more intense games. Arguably the peak of NES graphical design came near the end of the NES’ lifespan when a young developer, Masahiro Sakurai, released Kirby’s Adventure, weighing in at a whopping six megabits (750 kilobytes) and pushed the aging console to its limits.
[The Ricoh 6502] also succeeded in attracting the talents of a young programmer, Satoru Iwata…
For all its potential, the NES had some critical limits. A relatively slow processor and hard sprite limit of eight meant that some game genres (such as shmups–Summer Carnival ‘92 Recca notwithstanding) were very difficult to pull off without slowdown. Color, sound, and controller limitations all held back what was capable on Nintendo’s first successful console. Despite its limitations, however, the NES dominated the marketplace, selling over sixty-two million units worldwide and shipping over half-a-billion units in software sales, taking Nintendo to the top of the games industry in the process. For a console that leveraged such outdated technology, the NES’s impact was incredible catapulting Nintendo to the heights of its popularity.
Super Nintendo Entertainment System* (SNES): The Aging Prizefighter
Despite riding on the NES’ popularity, the SNES nevertheless marked a notable jump in performance from the aging NES. Packing sixty-four times the work RAM and a 16 bit custom Ricoh 65C816 (also known by its chip ID, the 5A22) clocked at 21.47727 MHz (21.28137 PAL), the SNES provided a much needed boost for game developers. Suddenly, genres that had struggled on the NES, like shmups and RPGs, were no longer held back by hardware limitations and saw an incredible boost in gameplay quality. Even other genres like platformers, which had ran fine on the NES, saw an enormous increase in visual fidelity, allowing for parallax scrolling and more impressive sprite work among other features.
Facing off against its biggest competitor, the Sega Genesis, in the “console wars” of the 1990s, the SNES’ major advantage didn’t come from graphics, but from its sound.
For all it’s technical wizardry, the SNES suffered from what games historian Dominic Arsenault calls a “decentralized architecture.” Following in the footsteps of famed Nintendo developer Gunpei Yokoi’s motto of “lateral thinking with seasoned technology,” SNES designer Masayuki Uemura focused on surrounding the relatively weak 65C816 with powerful components that “cluttered and complicated the programming process.” Similarly, Nintendo’s claims of 21 MHz represented a best-case scenario. Most of the time, the SNES operated at “an effective operating speed […of…] 1.79 to 3.58 MHz.” Such a slow CPU (at least when compared to the available options on the market at the time), meant that developers needed to harness all of the SNES’ various “specialized components” in order to produce quality games.
Facing off against its biggest competitor, the Sega Genesis, in the “console wars” of the 1990s, the SNES’ major advantage didn’t come from graphics, but from its sound. Rocking a custom Sony audio chip (designed by none other than the legendary Ken Kutaragi) as one of its central components, the SNES allowed composers to create much more varied soundtracks than the comparably stunted NES. Games like Chrono Trigger, Donkey Kong Country, Final Fantasy VI (III in the United States), F-Zero, and Super Mario World pushed the SNES’ sound chip hard, and it showed.
Like the NES, the SNES’ adaptability was one of its greatest strengths. Add-on technology like the DSP-1 (PilotWings and Super Mario Kart), SA-1 (Super Mario RPG and Kirby Super Star), CX4 (Mega Man X2), and a myriad of others, augmented the system in ways not possible before, allowing for experiences that were faster and better looking than on the base console. Still, Nintendo pushed for more. While the SNES’ Mode 7 allowed for backgrounds to be manipulated, stretched, and flipped around in order to simulate a 3D environment, but the base SNES was still incapable of true 3D gameplay. Nintendo wanted to change that.
To address it, Nintendo assigned the development of a SNES add-on chip to Argonaut Software, an upstart British programming team, to develop a 3D capable chip for the SNES, after seeing their spectacular work on the Game Boy game X. The result was the Super FX Chip, an incredible piece of engineering that introduced groundbreaking polygonal 3D gameplay for the first time on a Nintendo console. Both the Super FX chip, used in the original Star Fox (Starwing in PAL regions), and its successor, the Super FX 2, used in Super Mario World 2: Yoshi’s Island and Doom, were powerful chips that came too late in the SNES’ life to make a substantial difference. By that time, Nintendo was already competing against the Sony Playstation, a revolutionary new system capable of much more complex 3D output than what the SNES could muster.
The SNES, ultimately, would prove to be a financial success, just like the NES before it, selling 49 million consoles and 379 million software units before being discontinued in 2003. While some, like Arsenault, see it as the beginning of the end of Nintendo’s iron grip over the gaming industry, others see it as the apex of Nintendo’s creative genius, a golden age of game design, and the perfection of sprite-based graphics.
*All citations, unless otherwise noted, are from Dominic Arsenault’s Super Power, Spoony Bards, and Silverware: The Super Nintendo Entertainment System (Platform Studies) published by The MIT Press. You can find it here.
Nintendo 64 (N64): The Crippled Giant
That would change with the next generation, with the ever-divisive Nintendo 64. As the company’s first 64 bit, 3D focused console, the N64 took Silicon Graphics impressive MIPS architecture and introduced it to the world of video games. Operating its 64 bit NEC VR4300 CPU at 93.75 MHz and its famed Reality Coprocessor at 62.5 MHz, the N64 allowed for impressive 3D rendering that theoretically allowed the console to outperform Sony’s nearly two-year-old PlayStation.
Some of the N64’s hardware gave it an advantage over the competition. Built in trilinear mipmapping allowed for cleaner textures than the competing PS1 and Saturn, edge-based anti-aliasing paved the way for today’s FXAA and MLAA equivalents, real-time lighting was revolutionary, and its fully-programmable GPU allowed some developers to harness its true potential through tweaking microcode. Similarly developers on the N64 pioneered several techniques still in use in modern, 3D game design, including level of detail scaling, clipping, large environmental textures, and advanced texture streaming. Indeed, for developers talented (and stubborn) enough to work with the N64, great things could be achieved.
Unfortunately, flaws in the N64’s design crippled its potential. With only 4 KB of texture memory, developers were forced (at least until late in the hardware’s lifetime), to make “serious concessions in texture design.” They faced either smearing small textures across large surfaces or using Gouraud shading to make up for the console’s insane limitations. While this was a decent workaround, it resulted in a notably cartoony look that didn’t match up to the “more realistic look of competing PlayStation games.” Additionally, a flaw in hardware design made the Reality Coprocessor, the N64’s GPU, go through the CPU to access memory, an inefficient design decision that crippled the N64’s polygonal output to a tenth of what was theoretically possible.
Nintendo’s decision to adopt cartridges, however, was perhaps the biggest blow to the N64’s technical prowess. Frightened by piracy, Nintendo stuck with the cartridge, a storage medium whose range of 4-64 MB left much to be desired when compared to Sony’s 650 MB CD. Where PlayStation games like Final Fantasy VII had entire FMV cutscenes spread across multiple disks, N64 games had to settle for real-time cutscenes instead. While cartridges were much faster at loading, the vast disparity in storage capacity coupled with increased cost made cartridges yet another roadblock for the N64’s popularity.
for developers talented (and stubborn) enough to work with the N64, great things could be achieved.
Despite its many successes, the N64 failed to capture the magic of either the NES or SNES, selling 32 million consoles to the PlayStation’s 102 million. It sent Nintendo home empty-handed, rewarding Sony’s innovation and consumer communication and punishing Nintendo’s stubborn reticence. After spending most of the late 90s in a desperate attempt to catch up with Sony, Nintendo decided that it had to take a few chances and compete with Sony directly on power…
Be sure to check out the second part of this series where we discuss the technical history behind both the GameCube and Wii.