Silicon Labs began its evolution towards the Internet of Things (IoT) in 2010, at which time Tyson Tuttle, who started with the company as a design engineer, was CTO. Since taking over as CEO in 2012, Tuttle has grown that early IoT investment into half of Silicon Labs’ nearly $650 million in annual revenue, and has his sights set on increasing it by driving simplicity (ironically, through software) deeper into the IoT marketplace.
A 2016 Top Embedded Innovator, Tyson Tuttle, CEO of Silicon Labs, discusses system on chip (SoC) integration for the IoT, the need to abstract complexity from embedded development, and what it’s like to lead a company from the ground up.
How are SoC architectures evolving as we enter IoT era?
TUTTLE: If you look at SoCs in a cell phone, you have a big digital baseband chip, you have an LTE chip, then you’ve got separate power management, you’ve got separate radios, power amplifiers, the memory, the RAM, and the flash – all of those are a separate die. In IoT, all of those are getting integrated into a single die. So you can think of the cost point for mobile phone-type of applications being anywhere from $20 to $100 worth of electronics. In the IoT, that’s perhaps a factor of 10 lower because the volumes are so much bigger but the integration level is higher. Maybe the applications are simpler, but that drives single-chip implementations.
The SoCs in a cell phone are high-powered quad-core or eight-core ARM processors, and in the IoT it might be a couple of cores worth of smaller scale ARM processors. But it’s not just about processing power. It’s about the integration of the connectivity, of the power management, of the sensor interfaces, of the memory, etc. So that leads you to a different process technology because you’re not in a digital-only process – you have flash memory integrated in with logic, in with RF, and mixed-signal interfaces as well. So it is a different type of SoC for IoT compared to computing or mobile phones.
I would also add that the power levels are substantially different. In a mobile phone you’ve got a 3 amp-hour battery that lasts a day (on a good day), whereas in the IoT you may have a 200 or 300 milliamp-hour battery. So you have one-tenth the battery capacity, but that needs to last 5 to 10 years. You’ve got orders of magnitude difference in terms of available energy, and that fundamentally means that you’re going to need to be sleeping most of the time – your chip’s not going to be doing full-scale processing and communicating 24/7. You’ve got different energy requirements and constraints, which also leads to different optimization of the SoCs. They need to be able to hibernate and use very, very little power, and even when they’re powered up they need to be able to use as little energy as possible.
For a company named Silicon Labs, how is the IoT driving your investment in software and middleware? How do you see the Silicon Labs portfolio evolving over the next 5-10 years?
TUTTLE: The silicon is the foundation of the house. It is the table stakes, and ultimately that is going to be the heart of IoT devices. But then the software, stacks, and middleware are becoming a much, much bigger part of the product. You have to have a communication protocol, you have to have development tools, you have to have applications that run on those, you have to know how to communicate, and all of that is defined by software, middleware, and development tools.
Over the next 5 to 10 years the software component of our products will continue to get more and more sophisticated, so you can have a common piece of hardware that then proliferates into a lot of different versions of products that are targeting all these different applications. And, when you’re talking about IoT, you’re talking about thousands of different applications. So, you will have a few different varieties of hardware and different levels of optimization, but it’s much more granular on the software side. We view that becoming very, very critical. If you look at the last three or four years, most of the investments in R&D and acquisitions have been, essentially, software. So, in some ways, the silicon is not the hard part.
The biggest challenges, and another aspect that’s driving our investments, is this whole concept of simplicity. If you think about thousands of applications and tens of thousands of customers, being able to abstract the complexity to make it easy for customers to design in, easy for us to support, and things as simple as possible, is actually a really difficult engineering problem. But it’s really the key to scaling the business in these broad markets. You can only handhold so many customers, so making things easy (a lot of which is software, some of which is also modules), building that support ecosystem, and driving simplicity is a key piece of our investment strategy in IoT over that timeframe.
You worked your way up through the ranks at Silicon Labs to eventually become CEO. How do you think that path differs from CEOs that, for better or worse, are hired from the outside?
TUTTLE: I’ve been with the company 19 years and started as a design engineer and worked my way up through product marketing. I started a couple of the businesses and moved into the general manager ranks. I was CTO for a while. I was COO for a while. At a technical level, a business level, and with knowledge of the teams, I’ve helped craft the strategy. I was not a founder, but I was the tenth employee, and have been involved in pretty much everything since day one. That is different than somebody coming from the outside.
I also always try to learn from experience and from people with knowledge from other companies who’ve had other experiences. Through the acquisitions we’ve made we’ve hired a lot of extremely talented people over the years, and being able to pull all that together into cohesive strategy and cohesive culture is something that has been a 19 year journey. Certainly people from the outside can come in and do a great job. Necip Sayiner, our prior CEO, came in from Bell Labs and did a great job.
So I think it can work both ways – each personality is different and each company is different – but given the culture and given my path, I think I’ve been able to steer the strategy in a way that builds consensus and builds enthusiasm that’s different than if somebody came in from the outside and dictated. It was a little bit more from the bottom up than driven from the top down.