There are several lessons the data center industry can learn from Nintendo’s Switch, the company’s history, and its design philosophy. Drawing parallels to the video game industry might seem like a stretch; like perhaps I’ve been playing Dead Cells so much over the last week that my brain has turned to mush. While the latter is partially true, stick around for a strategic discussion – albeit one off the beaten path.
Lesson 1: Lateral Thinking Through Weathered Technology
The inventor of the Gameboy, Gunpei Yokoi, had a unique perspective and design philosophy. One day, Yokoi saw a bored commuter on the train playing with a calculator to entertain himself. It birthed the idea of the Game & Watch LCD handhelds. When it came time to evolve the concept into the Gameboy, Yokoi decided to use older, tested and trusted technology rather than the bleeding edge.
Sometimes this philosophy is translated as “withered technology” but that’s too pejorative and misleading. The lesson here is that lateral thinking about established technologies is usually what triggers evolution. A good example is RFID tags. RFID tags have been around for decades and are unquestionably weathered tech. However, the Internet of Things means they’re being used in new, interesting and creative ways. Cloud computing was built on commodity hardware. Virtualization reinvented weathered server technology. Containers went a step further in regards to virtualization.
The lateral thinking piece is key – finding new and interesting ways to use established technology is what often births new, exciting tech.
Lesson 2: ARM Processors
Let’s look at the Switch’s internals. The system leverages ARM processors.
How ARM processors apply and effect the data center industry has long been a hot topic of discussion. I’m not going to get into CPU architecture, but here’s a quick primer:
Intel x86 architecture-based processors are generally used in PCs and servers, while ARM rose to prominence in mobile devices. The primary reason ARM chips are appealing for mobile devices is because they have low power draw compared to x86. ARM was integral in enabling phones to become powerful pocket computers while retaining decent battery life.
I had a Sega Game Gear when I was a kid – that thing mowed through 6 AA batteries every 2-3 hours. I know the pain of power-thirsty devices.
Data centers consume massive amounts of power, and it’s a constant struggle to lower the cooling bill. This is why ARM is enticing in the data center. While ARM might be considered nascent in the data center world, it’s a weathered technology in mobile devices. The lateral thinking comes in applying ARM in new and interesting ways. Intel recognizes this, and has focused on rearchitecting x86 to compete with ARM’s low power draw, while ARM races to compete with x86 performance.
The Switch is a perfect example of ARM’s crossover appeal. Nintendo designed a system that is both a handheld and a console. This meant creating a handheld device with the power to provide a console experience. Anyone who grew up with an NES and a Gameboy knows how big the delta between these two experiences used to be. Playing handheld versions of their console counterpart games meant major concessions.
ARM’s low power draw enabled handheld and console worlds to converge. The Switch is a mobile device powerful enough to act like a console. A hybrid. Some hardcore gamers argue the Switch isn’t powerful enough to handle some of the latest, greatest games because of its ARM architecture – this is in line with thinking ARM isn’t suitable in the data center.
Like the Switch, ARM adoption in the data center isn’t meant to compete and replace x86. It’s meant to redefine and create an all new category of IT architecture. The move towards smaller, distributed deployments at the last mile in conjunction with hyperscale is rewriting how IT architecture is done. There’s a massive opportunity for ARM (or low power chips in general) in this evolution. It’s why Intel is investing big in re-architecting x86 chips for lower power draw to compete.
The first company to unite handheld and console experiences was technically NEC. The Turbo Express played the same games as the Turbo-Grafx-16 – but the bleeding edge is expensive. The unit was $300, atop the $150 for the console. NEC was also first to market with a CD-based console, even beating Sega. Being first isn’t always best.
I can’t stress how massive a risk Nintendo took with the Switch. But I’ll try.