Tomorrow Tech: Hardware vs. Software

In 1965, Gordon Moore, then the head of research and development at Fairchild Semiconductor (he later co-founded Intel), posited that “the number of transistors in a dense, integrated circuit doubles about every two years.” In layman’s terms, this meant that processing power was increasing rapidly, allowing the computing industry to set the stage for today’s lightning-fast processors, phones and tablets. But how long can Moore’s Law of predictably big gains hold true?

While some experts feel that Moore’s Law sunsetted sometime between 2010 and 2021, others believe that it’s still in play but that we’re also experiencing Moore’s Second Law (aka Rock’s Law), namely that the capital costs of semiconductor plants increase exponentially over time.

So, what does this mean for boaters accustomed to buying significantly faster multifunction displays every few years? Have we reached a plateau phase of chip and hardware maturation, where the biggest gains that customers can expect in coming years will come from ones and zeros, not silicone?

“While mobile phone [technology] is leveling, I don’t see that happening in the marine market because we’re a little bit behind the bigger computer-electronics market, and because of the new sensor technology that we can bring in,” says Eric Kunz, Furuno’s senior product manager, adding that AI could also be a coming game-changer.

Dave Dunn, Garmin’s director of sales and marketing for marine, points toward the confluence of Moore’s First and Second laws coupled with basic personal economics. “There are chips out there that can handle more than the marine market will ever need, but they’re very expensive,” he says, adding that most users will never max out current-generation chipsets. “There’s only so much [processing power] that you can see with the naked eye.”

Another factor that’s hindering all sectors can be loosely termed “COVID’s Law”—namely that when a pandemic wreaks havoc on supply chains, component shortages can stymie hardware’s evolutionary curve.

“Manufacturers are still pushing forward with innovations,” Dunn says, pointing toward today’s shortages. “[At Garmin] we’re focused on what we can build today, and we’re building [hardware] with longer life cycles. Hardware isn’t quite as important as software, but it’s pretty important.”

This, of course, means that if the silicone will be relatively static, the ones and zeros need to be more than just efficient and innovative: They need to dazzle.

“The black magic is always in the software,” Dunn says. “We’ve got an in-house joke: ‘It’s just software; it can do anything. So how hard is it?’”

Kunz agrees: “We’re always talking about optimization, and how we can take our existing code and optimize it to run faster. There’s more [performance] that can be squeezed out. You can go back and take something that doesn’t work well and recode it.”

This can be especially important with networked third-party products. “If there’s a [third-party] product on the network that works through the MFD, it can bog down the system,” Dunn says. “So much can be integrated, but if it’s not done right, it can create a poor experience.”

One tried-and-true solution involves throwing more processing power into MFDs or black-box processors during manufacturing, but—given the scarcity and cost of today’s chipsets—ones and zeros can often deliver a more cost-effective solution. Take Garmin’s OneHelm vessel-monitoring and control system, for example. “With OneHelm, we work with other [manufacturers] to make sure their code works [on our operating system],” Dunn says.

Even when third-party product relationships aren’t involved, Kunz says that there’s always a balancing act between developing existing products and platforms and working to create the next generational step change (one that’s likely facilitated by next-generation hardware). “There’s a time-value relationship between developing costs and what customers get out of it,” he says, adding that too many user-interface (UI) changes can be problematic. “It can screw up customers,” he says, referring to changes that require customers to relearn how to interact with their glass.

But, as we have seen with mobile devices and MFDs, evolutionary UI improvements—if leveraged carefully—allow manufacturers to refresh their customer experience without leaving some longtime users searching for basics such as a web browser’s address bar (a recent Apple UI change).

So, as we approach (or potentially surpass) the crossroads of Moore’s First and Second laws and COVID’s Law, customers can expect constantly improving ones and zeros that are periodically advanced by better, albeit likely more expensive, chipsets and bigger, brighter, and sharper-looking glass. Which, when one ponders the lessons from high school economics class, could be a good thing.