Home5GEthernet inventor on IoT, innovation, more: Bob Metcalfe Q&A

Ethernet inventor on IoT, innovation, more: Bob Metcalfe Q&A

Editors Note: Bob Metcalfe invented Ethernet in 1973 while working for Xerox in Palo Alto, Calif., and went on to serve in numerous roles with 3Com, InfoWorld and Polaris Venture Partners, among others. He is also Professor of Innovation in the Cockrell School of Engineering at The University of Texas-Austin, and, this year, is serving as a Visiting Innovation Fellow at the Massachusetts Institute of Technology. 

Metcalfe joined RCR Wireless News Managing Editor Sean Kinney to take a look back on his creation of Ethernet as a way to connect Xerox workstations to the still-burgeoning Internet, and discuss the transformative impact the technology has had on the way people around the world conduct their personal and business lives. 

Sean Kinney: Bob, you invented Ethernet in 1973 while you were working with Xerox, and went on to work in a number of networking technology venture capital companies, and you currently serve as a Professor of Innovation right here at UT Austin. I think you’re also, this semester, a visiting Innovation Fellow at MIT where you did your undergrad work. So, let’s just jump right into it, Bob, and start – and I hate to make you look back and walk us down memory lane, but tell us about, in 1973, when you and David Boggs invented Ethernet – what was the problem you needed a solution for, and how did you go through the research and development that led you to Ethernet?

Bob MetcalfeDave and I were among the luckiest people on Earth, because we were given an important problem that no one had ever had before. The problem was: How would you connect a building full of personal computers, one on every desk? Can you imagine that – a computer on every desk? Because in 1973, no one could imagine a computer on every desk. And then we knew we were building arguably the first laser printer, which ran a page per second – 500 dots per inch, 20 megabits per second. The existing network technologies were way too slow to keep that printer busy.

And there was an Internet around. It was called the ARPANET then, and everyone wanted to be connected to that. So we began designing a network that would connect a building full of personal computers. That was good luck – getting a great problem.

Kinney: You mentioned ARPANET there. And I think another one of the predecessors to Ethernet was ALOHAnet there at the University of Hawaii. What’s the key differentiator from those two systems when you compare them to what you and Dave came up with?

Metcalfe: Those are the two foundational technologies for Ethernet, the ARPANET packet switching and the ALOHA network, which was channel sharing – a very clever way of sharing a channel. So we combined those two to come up with Ethernet. But neither one was a [local area network]. The ARPANET was not a LAN, and the ALOHA network was not a LAN.

And Ethernet was very much a LAN. So the differences mostly stemmed from that difference. ALOHA was packet radio in 1970, and now we enjoy Wi-Fi. People frequently ask why we did this cable stuff when we could’ve done packet radio in 1973. Why did you bring these cables in? All you need to see is a picture of an ALOHAnet modem. It’s about as big as a microwave oven, cost tens of thousands of dollars and ran at 4,800 bits per second.

We needed to run at hundreds of kilobits, if not megabits per second. And we didn’t have to go far the way that radio did. So it made sense in 1973 to make a copy of the ether. There’s the big ether out there, but we made a copy of it inside of a coaxial cable in order to get low cost and very high speed.

Kinney: I think IEEE last year observed the 40th anniversary of the Ethernet standard. Is that 40? I think that’s right – 40 or 41.

But looking back over those four decades, how would you summarize the transformative effect Ethernet has had on the way that business gets done around the world?

Metcalfe: That’s too big a question this early in our discussion. Ethernet has evolved rapidly over these 40-some years, as the world around it has evolved. And it’s been, as always, co-evolutionary – so the Ethernet that we have today bears little or no resemblance to the Ethernet Dave and I built in 1973.

The biggest competitor of Ethernet in its very early days – to get to your point – we used to send memos. Ethernet was invented in a memo, typed with a typewriter on a piece of paper, Xerox copied, then dropped off in inboxes around the building. So, Ethernet’s earliest competitor – and by the way, Ethernet cost thousands of dollars per card … to enable your personal, well no one had personal computers, but if you did, it would [have] cost you thousands of dollars to put it on Ethernet.

So, our biggest competitor was called sneakernet. People would say, well, why would I want to spend $1,000 on an Ethernet connection when I could just take my document, put it on a diskette, carry it – wearing sneakers – down the hall to the printer, and have it printed out. Why would I want to have cable installed and everything? So, sneakernet was our first big competitor.

Kinney: Looking back, I know you and Dave have different dates that you attribute to the birth of the Ethernet, but that aside …

Metcalfe: No, we don’t.

KinneyNo?

Metcalfe: No, we’re talking about different events. I’m talking about the day – May 22, 1973 – the date of the memo, and Dave is fond of celebrating the date it sent its first packet, which took some months to arrange, but they’re very close.

Kinney: But at what point did you two know that you had done something that would really pave the way for IT for years and years?

Metcalfe: There was no single time. Along those lines, there was an early event where – in the environment at Xerox Park – there was resistance to Ethernet. So in this new personal computer that was being built there – the Alto – as a researcher they’d give you a piece of paper, and you’d write I want an Alto. And then there would be a list of configuration options, one of them was Ethernet, [selected] by checking a box, Dave and I would build another card and put it in their Alto.

So, Ethernet was an option. And in fact, one of David Boggs’ most famous statements is: “Ethernet is not an option.” What happened was, we had the first 10 personal computers connected in Ethernet, and one day somebody took the terminator off the end of the coaxial cable, rendering it useless. Suddenly, 10 people stood up in the cubicle area and were looking around – what went wrong?

And it was in that moment that we knew Ethernet was not an option. Everyone had learned to use it constantly.

Kinney: You touched on this very briefly earlier in our conversation, but why Ethernet and then …

Metcalfe: The name, you mean?

Kinney: Yes, the name. And then a second question: why yellow?

Metcalfe: We chose half-inch-thick coaxial cable as the initial shared medium for Ethernet. But we knew in our hearts that different media – ultimately, the radio from ALOHAnet – would be used instead of this thick coax. So, we didn’t call it coaxnet, but we needed to think of something media-independent.

And there was this word floating around in physics – the ether, the luminiferous ether, which was, at one point in the history of physics, an omnipresent passive medium for the propagation of electromagnetic waves. It turns out the historical ether was proven to not exist, so we adopted the word – an omnipresent passive medium for the propagation of electromagnetic waves, which, in our case, would be data packets.

That’s why it was called Ethernet and not coaxnet, or twisted-pair net, or radionet, or ALOHAnet. The ether was a stab at media independence, which was vindicated. We went from half-inch, to quarter-inch, to twisted-pair, to optical fibers and, lo and behold, in the ’90s, we went to radio; now the yellow cable.

So here we are, building Ethernet, designing it and trying to put it in the building. And many people said no one is going to install new cables for a data network. We already have cables, we have telephone cables and we have power lines, so you’ve got to make Ethernet work on power lines. And we looked in the ceiling, and there was telephone cable and there was power cable, and they were all blue and gray.

We needed a color so that we could see the Ethernet cables in the ceiling clearly. It was a little bit of marketing, too, because at the early computer conferences where we demonstrated Ethernet interoperability – the Interop conference, which was created for this purpose – we had the cable be bright and yellow so it would stand out among all the other cables in the ceiling, which were generally black and gray, and boring. That’s why we ended up with yellow.

Yellow – I believe it’s the standard color for Ethernet cables, but there’s a hell of a lot of blue Ethernet cables around.

Kinney: And gray, apparently, too.

Metcalfe: And gray. And so let a thousand flowers bloom, but if you want to do it right, you’ll make your cables yellow.

Kinney: Now, if we could shift gears a little bit, I mentioned in the introduction that at UT, you are a professor of innovation. You and I had the opportunity to chat a little bit before we sat down here, and I’m curious about how you teach innovation. That seems like you have to go off textbook, certainly, in communicating that to students. But how do you quantify, and explain and repeat innovation?

Metcalfe: When I went out promoting the idea that various universities needed a professor of innovation, that was the first question: “What is innovation?” – which leads to a semantics argument that lasts forever. And then even if you know what it is, how do you teach it?

So, to me, just to clarify my semantics, invention is what we did in the lab at Xerox to invent Ethernet. Innovation is what we did at 3Com Corp. to commercialize Ethernet, which is to make products and sell them competitively in a marketplace. Fight the standards battle.

So, I wrote a paper once titled “Invention is a Flower; Innovation is a Weed.” And you learn this as soon as you bring your innovations out into the real world. You find that people don’t want to be innovated upon; they’d just as soon stick with what they’ve got. And then you have competitors, so it becomes very rough-and-tumble. And you become a weed, and people are constantly trying to kill you.

So, when I started promoting my new career after venture capitalist, I settled on the term “Professor of Innovation” as being what I wanted to do, because I wanted to get involved in the teaching and learning – not only the invention part, but the harder part that comes later, when you have to engage the status quo. And the status quo is mean and resourceful, so I tried to embody that in the word innovation – bigger than invent.

And it includes entrepreneurship. Some entrepreneurship is not innovative, and some innovations are not entrepreneurial, so they’re not the same word. And I chose innovation, which, to me, includes big parts of entrepreneurship.

Kinney: When we talked earlier, I asked you for a good example of a place where innovation is happening and all the right ingredients are in place. You said Palo Alto. So, what are some of those variables that are in place in Palo Alto that could potentially be re-created elsewhere in the hopes of fostering that same kind of innovative ecosystem?

Metcalfe: I lived and succeeded around Palo Alto for 23 years, so you’ll never catch me speaking ill of Palo Alto. But Palo Alto isn’t the only place where innovation occurs. And most research universities are fruitful areas. In fact, they’re my favorite kind. Startups out of research universities [are what] I consider to be my specialty, what I’m really interested in.

Silicon Valley is, to me, the gold standard, as you put it earlier. And so around here and Austin, I’d say my mission is to help Austin be a better Silicon Valley. Once again, there’s semantics in that, too. There’s a good deal of Austinites, including those especially here in East Austin, who don’t like it when I say I want Austin to be a better Silicon Valley. They would prefer it be weird – keep Austin weird.

I’m not buying into this weird stuff. San Francisco would kill Austin in a weird contest. So that just proves you can be weird and you can be Silicon Valley; you don’t have to choose.

Kinney: I have a follow-up question to that, but for a gauntlet from Bob Metcalfe – San Francisco’s weirder than Austin; do what you will with that challenge – but …

Metcalfe: Oh, come on. There’s no question.

Kinney: Aside from a research university, what are some of the other factors? We talked a little bit about just availability of high-quality, high-speed Internet. Surely that goes a long way toward helping these tech-centric ventures see what they can do, you can’t be limited by that.

Metcalfe: In this world, having Internet is a big part of the infrastructure for innovation, which is why I say, keep Austin wired – same letters; just rearranged. The proliferation of the gigabit Internet – in the ’70s, we had the kilobit Internet. Then we went through the megabit Internet. Now we’re going to the gigabit Internet.

And that’s an important part of building infrastructure, and Austin’s right on that. We have three, or four, or five different ways to get gigabit Internet now, led by Google, [which] provoked it with Google Fiber. Getting gigabit Internet in your city is a great way to enhance its participation in innovation, because you can connect to other places, get critical mass, share information and then even build systems on top of it.

Kinney: That is a huge part of what we cover here, but the wireless end of it doesn’t necessarily work unless there’s a robust wired infrastructure. You mentioned Google Fiber. Google has done a great job with marketing and elevating that sort of presence of mind in the general public. And here in Austin, we have AT&T doing the same thing, Grande Communications, several others. And we’re seeing that all over the country, too.

What do you think is driving that trend? I’d like to say I think these service providers are trying to open up innovation to the towns and cities across the country, but I don’t really think that’s the case. What do you think is driving it, and what do you think some of the potential obstacles are here?  I’m particularly thinking of [Federal Communications Commission]-type regulation …

Metcalfe: Oh, you beat me to it. I was going to say the FCC is the principal potential obstacle here. It takes seriously making the Internet into a utility. That does not bode well for the proliferation of the gigabit Internet. One of the reasons the carriers and their suppliers are putting out gigabit fiber is the continuation of a long trend. It’s been true since we came out with 2.94 megabit-per-second Ethernet in 1973.

The Ethernet and Internet world has been a build-it-and-they-will-come world. So if you up the bandwidth, out come all these new applications that were unanticipated. I think a lot of the investment is driven by that, but there’s also a new anchor tenant. One of the things that drives gigabit anything is television. The evolved television with all that video – most of the traffic on the Internet is video now, and it’s going up, and up, and up, and up, and up. And that’s the anchor tenant who drags along all that infrastructure and gives us the benefit of the gigabit Internet.

Kinney: If we could circle back to innovation – and I enjoy this sort-of using Austin as an example, because we both, to varying degrees, understand what’s happening here. We’ve got the research university, we’ve got high-quality internet. What other things can be done here in Austin to sort of boost this startup tech ecosystem we have?

Metcalfe: I use the Silicon Valley model. And if you go to Silicon Valley, there’s a couple of really fancy research universities there, like UT and [Texas A&M University] is not that far away – so not just one university. But we also have a tradition of entrepreneurship. The culture of entrepreneurship is strong in Austin, going back several decades, [while] not as far back as Silicon Valley, but pretty far back.

That leads to accelerators. It leads to people. So, after you have the idea for a company, you need to recruit people to make it happen. And a lot of those people are here, waiting to be recruited. The brain trust, the people who have started several companies, are extremely valuable to starting your company because you can go recruit them. And that’s a big part of Silicon Valley – all that human capital sitting on the shelf, waiting to be scooped up.

Then there’s venture capital, and we have some venture capital here. There’s a big argument now whether we have enough venture capital in Austin. I think we have plenty. I think we just need to increase the quality of our companies, because venture capital knows how to fly. And the big VCs invest here already; they just need an excuse, and the excuse is a really cool company. So, I’m focusing on improving the quality of the companies.

And on the human capital point, I think the limiting factor in our startup culture now is qualified startup CEOs. If we had more of those, we could put more money to work getting better innovations from better companies.

Kinney: Does that assume that you can create an effective startup CEO – or I guess I was under the impression that you evolved into an effective startup CEO.

Bob Metcalfe: We agree; you evolve into it: You go to work for a startup, learn how to be an individual contributor, how to be a manager, how to be VP of this or that, and then someone taps you to be CEO. And then they tap you again to be CEO. It’s a lengthy development process. You can’t create a qualified CEO out of thin air, which is why most qualified CEOs are not 18 or 19 years old. Most of them are in their 30s and 40s. Although you often get the impression that all companies are started by 18-year-olds, the statistics don’t support that point of view.

Our undergraduate students love to rattle off Gates, Jobs, Dell … And then I ask them to name five more – that’s when they have trouble.

Kinney: You’re in that breath. You started a company out of a Palo Alto apartment, right?

Metcalfe: I did. I was 33 years old when I did it.

Kinney: You weren’t 18?

Metcalfe: No, I wasn’t 18. I finished college and went out and got a job in the world. And then eight years after my PhD, I started my company at the age of 33. In entrepreneurship, you tend to teach what worked for you. So, in the kind of instruction that I love to give, you’ll hear me advocating what I did. That’s fairly normal since people tend to advocate what worked for them.

Kinney: If we could change gears again, I really want to ask you about the “Internet of Things” or the Internet of everything, as I’ve heard it called. I think all of the projections suggest that 2020 is the big standout benchmark. We’ll have 50 billion connected devices and objects all communicating over various networks. And that’s an expansive conversation. So, let’s focus a little bit on the automotive vertical.

I was looking back at one of your Reddit ask me anything threads, and you posited that this would be a complex thing, because you’re really talking about three networks, right? One inside …

Metcalfe: Or seven.

Kinney: Seven? So, it’s expanded since a year ago, but …

Metcalfe: Well, I mean lots of them.

Kinney: Just take us through how you see some of those connectivity requirements; car-to-car, car within itself, I guess even car communicating with transport infrastructure to some degree. How do you see that evolving and gaining adoption in the automotive vertical?

Metcalfe: As the various bits and pieces of the car gain intelligence, the existing intra-car CANs – car area networks – get more elaborate. I hesitate to mention this, but there’s a whole branch of Ethernet now called automotive Ethernet, aiming to make that car area network get the benefit of Ethernets.

My car already has Wi-Fi in it. And when I get inside, I have access to the Internet automatically, just getting into my car, through Wi-Fi, through LTE, through the cellular network. So, that’s a second network. And then there’s the GPS in my car that’s up in the sky. We have a network sending us location information. And there’s the radio network that cars listen to.

Then there’s the network that’s going to be needed between the car and the road, so the car will communicate with the road to keep track of the road and what’s on it. And then there’ll be the network between cars. Cars will be talking directly to each other to evade collisions, for example, or to get help, or just to chat. So, let’s see – what other networks might there be?

Kinney: Two more.

Metcalfe: Two more? I have to come up with two more?

Kinney: To get to the seven.

Metcalfe: OK, let me think.

Kinney: Regardless, it’s an incredibly complicated networking scenario. And the ultimate promise here is fully autonomous driving, something that I look forward to immensely. But you’ll never have to drive, you’ll never have to worry about gridlock and plan your life accordingly. How do you see this gaining adoption, though? It’s not like we’re going to flip the switch.

And even from a consumer perspective, it’s going to be a long runway for everybody on the road to purchase these vehicles. What do you see as some of the key benchmarks in adoption here being?

Metcalfe: What’s happening is, the movement to so-called driverless cars is happening faster than I imagined. Of course, a million people a year die in automobile accidents, so that’s a pretty strong pressure to fix that problem. They used to complain about the Windows user interface and how bad it was – and I would say as a joke, “Well, it’s not as bad as the car user interface – because you put your hands on the steering wheel, and if you go like ‘this’ at the wrong time, you kill your whole family. That’s a really bad user interface.”

Kinney: It’s high pressure.

Metcalfe: Yes. So we are highly motivated to get rid of this, because it’s killing people. In the United States, 35,000 people a year die in automobile accidents. So we’re motivated to solve that problem, and the technology seems to be delivering. They’re already testing them here in Austin, Google’s already running driverless cars around here.

And there are all sorts of complications. The lawyers are all saying, “Who will you sue when you’re in a driverless car and there’s an accident?” Well, I’m sure that’ll be worked out. I’m looking forward to never going to the DMV again – never needing a driver’s license and not needing license plates, because I won’t own any cars at that point. So, that could happen.

When you say it could happen, it’s hard for weekly, daily, or monthly magazines to cover that story, because it’s a decade-long story. It’s going to take 10 or 20 years for us to make appreciable progress toward this. But it’s coming and I look forward to it.

Kinney: Well, it seems we’ve struck a nerve with our audience, because we’re getting some questions relative to this. The first one is: From a security perspective, how could you address IoT vulnerabilities, in a sort of speculative sense? What are the security issues related to IOT networks, and what are some potential solutions?

Metcalfe: I’m not an expert in this area. My observation is that these security problems are perennial, and they get solved by degrees, and then a new one flares. I wrote, arguably, the first Internet memo in 1973 on the first security invasion of the Internet. And it didn’t go away. I wrote the memo intending to motivate people to solve the problem in 1973, and you perhaps have noticed that computer security remains a problem today. And I’m not an expert in it.

I just have confidence that we’re going to – it’s a never-ending struggle. People think home security is solved, because there’s a lock on my front door, so my home security is solved, isn’t it? Well, no. People could break the window. Somehow we’ve solved security by degrees, and gotten comfortable with the degree of solution that we have in the physical world. Now we need to develop an intuition about what to get comfortable with in the cyberworld for security. But experts tell me it’ll never be solved.

Kinney: Now a personal question that’s come in about you. It’s about IoT. I guess the folks at home see the smart watch on your wrist there, and they’re curious if …

Metcalfe: My Apple watch.

Kinney: Yes, they ask what model you’re using and how you like it; how do you use it?

Metcalfe: When Steve Jobs showed me the spreadsheet on an Apple II, not having been an accountant, I didn’t get it and was kind of slow to adopt. Then when Marc Andreessen showed me the Internet browser at University of Illinois, I didn’t really get it.

I now have this point of view that whenever something new comes out, I am the first one to buy it and try to figure it out – I’m still figuring this puppy out. It tells the time, it tells the temperature, and I can get my heart rate. I do a lot of that, I’ve been preoccupied recently with my heart rate for some reason. And yesterday, I did my first “Dick Tracy” telephone call. My wife called me when I was walking around the lake. And surprising to me, my left wrist started pulsing. And I hit accept, and then I’m talking to my wife through my wrist radio.

So I’m in the discovery phase of these watches. I have the big one, because I’m a big person. And I expect it’ll be evolving rapidly. I’m very hopeful. By the way, my heart rate is … it said, try again. So apparently, it detected no heart rate.

Kinney: Do I need to call somebody, or can we wrap-up here?

Metcalfe: No, I’ve made it this far.

Kinney: Next question from our viewers is: What role do phone companies – I assume they mean fixed-line and wireless phone companies – play in pushing broad innovation? What can the AT&Ts of the world do to innovate?

And I guess I have trouble with that, because from my perspective, at least, the bigger an organization gets, the harder it is to have that agility needed to innovate and then do something with it.

Metcalfe: I’ve long been hostile to the carriers, because in my day, when my attitudes were ossifying, the telecom was AT&T, and we had to break the then-AT&T monopoly to build the Internet, and we had to break the IBM monopoly to build the Internet.

So, I’m slow to overcome that early hostility to the carriers. But the carriers provide innovation in two ways. One is, they provide a platform; just by upping speeds and availability and mobility, whole new possibilities are created for new apps. I don’t have to list them; they’re all over the place. But the carriers also drive innovation to the construction of the infrastructure. It was, in fact, AT&T that introduced dense-wave division multiplexing on optical fibers in the mid-90s. And that’s the silver bullet that sort of makes everything today possible.

I look to the carriers operating under the free-enterprise system, pursuing revenue where they can, satisfying customers where they can, driving innovation out the front or back door.

Kinney: We have a few more questions that have come in, and I’d like to get through a few of them. And one of them is not a question, but I’m going to mention it, Bob, they love your Chuck Taylors. So, that’s an affirmation, if you needed it.

But more carrier-specific: 10 years from now, describe what you think a carrier will be in a business sense. What will AT&T be doing a decade from now?

Metcalfe: I have no idea.

Kinney: Then another innovation question: In a hypothetical …

Metcalfe: Excuse me; let me revise that last answer. I have a fear about it. The FCC has recently decided to go back to the model that we broke – the FCC monopoly model of telecommunications carriers. It has recently decided that it wants to go back to regulating the Internet as a utility, and that scares me. It’s a deterrent to the investment by carriers and disruptive to the innovation process.

One of the things the FCC has expressed hostility to is fast lanes, where some people can buy more bandwidth than others. And somehow, that’s bandwidth inequality, and we need network neutrality, blah, blah, blah. Well, if you don’t let the early adopters pay extra for fast lanes, and require everyone to have fast lanes all at the same time, that will deter innovation – because the way innovation works, as I’ve seen it, is a few people get the cool gadgets, and then the cool gadgets get cheaper, and cheaper, and cheaper, and then eventually they proliferate.

If you held back the Apple Watch until everybody on Earth – all 7 billion of us – could have one, there’d be no Apple Watch because no one could afford to do that. That’s a partial revision to my earlier, nondescript answer. I’m afraid for the carriers and what the FCC might do to them – and by virtue of damaging their prospects, damaging innovation’s prospects.

Kinney: And we have another innovation question, a great hypothetical. They want to know: If you could start an innovation-oriented company today, with all of the funding that you needed, what would you do? What would you focus on?

Metcalfe: I would do computational genomics – computational because that’s something I know a little bit about, and genomics because it has huge potential for improving the human condition. I probably would not start another Ethernet company, we have a lot of those now so we don’t need a new one.

Kinney: They enjoyed that answer very much. I love this real-time feedback.

Metcalfe: How do you know?

Kinney: I’ve got a panel over here that’s giving me all this, and here’s another one, which dovetails with something I was wondering about. I read a lot about distributed antenna systems and small cell antennas, things of that nature, and power over Ethernet is really picking up in that space. The folks at home want to know: What are your thoughts on power over Ethernet? What’s the promise of power over Ethernet?

Metcalfe: It’s hilarious. As I mentioned earlier, in the early days of Ethernet they said no one is going to install new cables. You’ve got to get Ethernet to run over the power line. And now, 40 years later, irony of ironies, we’re insisting that the power be carried over the Ethernet cables instead. And so there’s a general trend as the Internet of Things proliferates, and the devices get smaller and more disbursed, we have to worry more and more about powering them, and if we can afford to put batteries in all of them.

Carrying DC out to these remote leaves of the Internet is a coming thing, and power over Ethernet is a fairly obvious response to how you carry the data and the power over the same wires to these remote devices? It’s a great thing.

Kinney: This kind of circles back toward the beginning of our conversation. As we wrap-up, I asked you earlier to take a look back at how Ethernet has been transformative in the 40-some odd years since it was invented. Can you speculate and look down the road for us, and tell us how you think Ethernet will continue to evolve and be relevant, and continue to be important to innovation over the next few decades?

Metcalfe: Well, if you’ll forgive me for calling everything Ethernet – Ethernet’s going in many, many directions. It’s continuing in the LAN direction, just going faster and faster. We’re at gigabit. We’re at 100 gigabit. We’re going to 400 gigabit. We’re going to terabit. That progression continues, and the evolution of media is continuing.

But Ethernet has stopped being a LAN, for many purposes, and is now the principal plumbing of the WAN. So, the old telephone synchronous SONET system is being blown away by Ethernet equipment in the WAN, not in the LAN. And then we’re going to the Internet of Things, which is the network going – it’s not called Ethernet, but it’s a lot like Ethernet – down into the tiny things. And most of that has to be wireless.

And one of the most exciting evolutions has been the development of carrier Ethernet. So what developed over the decades was a chasm between the WAN and the LAN. … The LAN had faster and faster Ethernet. And then there were the carrier services that connected the WAN to the LAN, and they were T1 or other pitiful old systems. And now they’re being replaced by gigabit fibers and what’s sometimes called carrier Ethernet – another direction it’s going.

So Ethernet is breaking out. Let’s just say networking is going in all directions. And you’ve heard of Moore’s Law, which is a real law, unlike Metcalfe’s Law. I like to think that Metcalfe’s Law and Moore’s Law – what they have in common is they begin with M.

But they talk about Moore’s Law sort of tapering off around now. I don’t think so, but they talk about it. But Metcalfe’s Law is going great guns, because we’re learning that when you network things, they are more valuable – a lot more valuable.

Kinney: Remind me of Metcalfe’s Law – the value of a telecommunications network is equal to the …

Metcalfe: Grows as the square of the number of users or connected devices. It’s a visionary thing for quantifying the network effect. But it basically says, if you network things together, they’re a lot more valuable than if you don’t.

Kinney: Is the math holding up?

Metcalfe: I did a paper in IEEE “Computer,” December 2013, which took the first 10 years of Facebook and explained it mathematically with Metcalfe’s Law. But having the math hold up is not nearly as important as getting the general idea, which is that networking is a good thing, and it brings value. And here comes the Internet of Things.

Kinney: Thank you so much for coming in  and sharing your thoughts on networking, Ethernet and IoT. We really appreciate it, it was very insightful, and I think the folks at home enjoyed it, based on the response we got. And we appreciate the folks at home tuning in to this edition of our RCRatx Studio Lounge Speaker Series.

Be sure to follow RCRWireless.com for lots more in this type of programming, and I’d encourage you to visit our website for lots of news and multimedia content. And again, we appreciate you. Bob Metcalfe, thanks for coming in.

Metcalfe: Thanks for having me.

Internet of Things IoT 2G LPWA
Previous post
Ingenu is creating an IoT-specific machine network
Next post
New LTE-U, Wi-Fi coexistence tests show promise, but controversy continues