Nintendo has, since the Wii, been on a weird hardware refresh cycle out of sync with Sony and Microsoft, and for good reason.
In 2005/2006, as the Xbox 360 and PS3 launched, Nintendo’s entry into the fray was the small, underpowered Wii. Jokingly referred to by some gamers as “two Gamecubes taped together” due to its anemic hardware specifications, it vastly outperformed expectations and was the top-selling home console of the seventh generation, beating its competition by around 20 million units shipped.
But the Wii couldn’t do HD output and third parties could rarely bring ports over from the PS3 or Xbox 360, instead leaning on PS2 ports when they could or making bespoke adaptations for the Wii that allowed the reuse of some assets scaled appropriately while delivering what was otherwise a completely different, Wii-centered gaming experience. In 2012, Nintendo released the WiiU, a similarly anemic hardware platform that was at least capable of high-definition video output over HDMI, but the platform couldn’t really handle ports even still, and thus it ended up being sort of a dud without a clearly fun gameplay gimmick like the motion controls of the original Wii.
In 2016, Nintendo made what is, in hindsight, an incredibly smart decision, merging the handheld business they’d been cultivating since the 1980s and their home console business into a single platform – the Switch. The Switch is, to me, a fascinating piece of hardware, in that it is the refutation of the high-end console market paradigm and the exemplar of what Nintendo has really been doing since the Wii.
The Gamecube was the last time Nintendo launched a system that was at parity on hardware power with the contemporary Sony and Microsoft consoles – in some ways better, in some ways worse, but generally easy enough to port to that Nintendo saw third-party titles flowing in and having just a Gamecube in your home wasn’t drastically limiting the scope of games you could enjoy. Since then, with both iterations of Wii and the Switch, Nintendo has largely avoided the pressure of moving to a platform with closely matched power to the massive monoliths of Microsoft and Sony, instead selling systems built on a strong catalog of first party titles, Virtual Console for a wide library of beloved classics, and strong-ish indie developer support, with a lot of digital sales being games from small developers. On the Wii, this worked decently well – less so on the WiiU due to the lack of interest in the hardware itself and the gamepad gimmick, but the Switch has seen Nintendo doing exceptionally well and brought their financial results back to the levels they were during the heady heights of the Wii.
On the hardware side, the Switch is such a departure in general from the industry mainstream but also even Nintendo’s strategy in general. Sony and Microsoft have strong custom silicon, and work with AMD and other partners to deliver specialized, custom solutions – highly powerful system on chip designs, accompanying SSD technology that drives a next-generation experience, and custom chassis design and thermal solutions that work to make that fancy hardware stay cool under standard operational conditions in a home, even when that design is kinda bad and lets RAM chips almost overheat (sorry, PS5). Nintendo’s past consoles were like this – the N64 was a fully-custom implementation of SGI workstation technology and had Rambus RDRAM a full 5 years prior to the technology’s brief use as the memory of choice for the Pentium IV CPUs, the Gamecube used a custom-built IBM CPU and ATI GPU, a relationship which continued for both the Wii and WiiU, with each new hardware spec gaining higher-clocked and more feature-laden versions of these bespoke parts.
The Switch is…none of that, though. It’s system on a chip is a bog-standard Nvidia Tegra X1, with the extent of customizations being reductions in clock speed. The chips specifically used for the Switch and Switch Lite (T210 and T214 respectively) are also used in other devices, like Nvidia’s various iterations of the Shield TV set-top box and various versions of Nvidia’s automobile infotainment systems. The other hardware inside is all largely off-the-shelf stuff – standard WiFi and Bluetooth controllers, standard eMMC flash memory for storage, and most other parts are standard issue stuff. That isn’t to say the Switch isn’t cool or impressive – instead, that the system designers and engineers spent most of their time working on actual hardware design, figuring out how to take those stock components and push them together into a sleek casing that could be used for both home and on-the-go play. They certainly succeeded, at least to my estimation – the Switch sells like hotcakes even in the face of next-gen platforms from Sony and Microsoft, and while there are issues like Joycon drift, the overall design works pretty well.
On the hardware power front, well, the Tegra X1 is a fine enough chip, with Geforce GPU cores from the GTX 900 series and a decent quad-core ARM CPU, but it isn’t even up to the level of the launch PS4 and Xbox One.
The trick is…it doesn’t need to be, though.
Nintendo sells consoles for Nintendo games, with other developers largely being sideshows. That isn’t to say that Nintendo has completely shirked the need for third-party development, but for 3 generations now, the vast majority of titles sold for Nintendo platforms are made by Nintendo. Most of us know that buying one too – getting a Switch today isn’t an investment in Call of Duty, Assassin’s Creed, or some other annual release title with a huge budget and marketing machine behind it. It is an investment in 1-3 Zelda games, a similar number of Mario titles, an Animal Crossing, maybe a Metroid, a couple of Pokemon titles, and a handful of fun Nintendo universe fare – Mario Kart, Super Smash Brothers, Mario Party, and Mario and Sonic at *Current Olympics at Time Of Release*. There are some people for whom Nintendo consoles are their main platforms of choice, and those people might want to have the option of third-party ports that are popular, but at the same time, after the Wii, Nintendo has clearly de-prioritized pushing a system for other developers in favor of making the best Nintendo-developed game system possible – and it works.
On a technical level, the base idea of a Switch Pro feels about right. The Tegra X1 is a chip made by Nvidia, and so with a few generations of GPU upgrades, a modern Tegra with Turing or Ampere architecture could be drastically more powerful per core and also support DLSS. For those not in the know on the tech, DLSS is “deep learning super sampling.” Originally marketed as an anti-aliasing replacement by Nvidia, DLSS has since evolved into a much-better replacement for temporal anti-aliasing techniques – using data about scene movement to reconstruct an image at higher resolution than rendered at. Most rumors of the Switch Pro hinge on DLSS being used specifically – with an undocked “Pro” model continuing with a 720p resolution screen but then using DLSS when docked to generate a 4k resolution display image for output to a TV. This is a pretty cool concept because it fits hand-in-hand with Nintendo’s hardware philosophy over the last 15 years – no need for excessive power when less will do. In this case, they can literally make a system focused on building the best 720p output, and then offer that natively on the device display itself, or scale it up for those playing on a TV. The logical portions of the GPU responsible for DLSS computation, the tensor cores, have no other role in standard rendering, so the system would be able to simply shut them off undocked and then unlock them while docked and run at essentially the same performance, so no more docked/undocked framerate comparisons (in all likelihood).
On a more Nintendo level, though, the idea is interesting, because there really isn’t a need for a Switch Pro. The WiiU came out once the Wii train was starting to falter and every retirement home, school, and most households that wanted one had one. While Nintendo’s bet that most people wouldn’t care about graphical fidelity or framerates paid off in 2006, by 2011, it was starting to slow things down as fewer third-party titles came out and the first-party slate was fairly bare, with all the major releases made. Right now, the Switch is doing great, by all accounts. It continues to sell strong, it had arguably the breakout hit of 2020 in Animal Crossing New Horizons, and by merging with Nintendo’s traditional handheld business and obsoleting the 3DS lineup in the process, a lot of third parties have a good compromise position – if they would have made a 3DS approximation of a major franchise, they can release better versions of that kind of game on Switch instead, or when that doesn’t work, release a cloud-based game instead. Nintendo has a slate of 2021 titles that look pretty good, from Pokemon Legends: Arceus to the remastered Pokemon Brilliant Diamond and Shining Pearl, Zelda Skyward Sword remastered, New Pokemon Snap, and continued DLC support for Super Smash Brothers Ultimate, Pokemon Sword/Shield, and Animal Crossing New Horizons.
So why would Nintendo release a new console soon, anyways? In my opinion, surprisingly if you’ve read my Sidenote posts, I actually…don’t think they need to.
Why is that?
Well, I mean, if I just think of my desires, a higher-fidelity Switch is a cool idea on some level. A console using DLSS and other modern GPU tech is a really fascinating concept, especially one built specifically to leverage it in a strong way. At a technical level, some of the more recent Nintendo releases are beginning to stretch the bounds of the Switch, with the latest, Super Mario 3D World having a new mode, Bowser’s Fury, that attempts to run with an unlocked framerate and frequently dips and stutters. It isn’t awful, necessarily, but traditionally, the point at which Nintendo’s developers run up against optimization hurdles is the point that they make new hardware. From a marketing perspective, the Switch has sold near the volume that most consoles sell in their full lifetime – while at 80 million units worldwide currently, there is around 20 million or so left before they start to brush up against the point where most consoles top out, and 101 million worldwide is where the Wii hit a ceiling. If you’re Nintendo, there’s some value in a Pro from a marketing perspective – jumping on the 4k hype as the technology has matured, being able to reach it easily with a mix of higher-performance GPU cores and smart upscaling, offering enough power for PS4/Xbox One titles to start to come over, which may allow for some limited amount of current-gen crossover, at least before more games begin to lean heavily on the power of the PS5/Xbox Series X.
There could be other benefits too – using a more modern storage design could make Nintendo games like Breath of the Wild that much more engaging, and updated technology for controls could lead to some interesting advances if they continue with the Joycon concept. Calling it a Switch Pro, as the rumors and “reports” have implies a degree of backwards compatibility or simple cross-gen support, such that titles in-development for Switch could implement a Pro mode or other small tweaks and then support a more substantial upgrade for titles that start development on the Pro hardware.
But I keep coming back to the original point – at the end of the day, does a Switch Pro feel like a necessary upgrade?
As much as the tech nerd in me wants to say “yes!” the truth is…probably not.