Phil McKinney’s job as CEO of CableLabs is to lead the cable TV industry forward into new technologies that can help with the delivery and adoption of cable TV services.
The company is the research arm of the cable TV industry, and its job is to push new technologies such as Docsis 3.1, which enables 10-gigabit-per-second download speeds and 1-gigabit upload speeds for cable TV broadband delivered over coaxial cable. That will change to full duplex, or 10 gigabits-per-second in both directions, over time.
I frequently interviewed McKinney in his previous job, when he was vice president and chief technology officer for Hewlett-Packard’s personal systems group. He was responsible for strategic planning and research and development for HP’s PCs. Now he’s helping to figure out how cable can stay competitive around the world against competitive threats such as 5G wireless, which will bring cable-like speeds to wireless data.
I caught up with McKinney at CES 2017. Here’s an edited transcript of our conversation.
VB: How did you wind up at CableLabs?
Phil McKinney: When I retired out of HP, I thought I was retired for good. My book was coming out. My wife and I were planning on going to Nashville. About three weeks in I got a phone call from a headhunter, looking for a new CEO. They said, “Phil, this is the perfect job for you.” I wasn’t interested. I thought they were calling me for a reference for someone at HP to fill a recruiting role. I said, “What makes you think that?” They said, “It’s a CEO role.” Um, you know I’m an innovation guy? I’m not interested in being a CEO, where I have to manage P&Ls and do all that stuff. “But it’s a CEO role with no profit and loss responsibilities.” How do you get one of those?
When they told me it was CableLabs, I turned them down. Tony Werner, who’s the CTO at Comcast, and Mike LaJoie, who was CTO for Time Warner, persuaded me to come to New York. I met with them, with Neil Smit, who’s the CEO at Comcast Cable, and Brian Roberts, the chairman and CEO at Comcast. They persuaded me to at least consider it. I spent five months in conversations with the CEOs. They were focused on transforming CableLabs along an innovation agenda. It had fallen into a very tactical, almost staff augmentation, very applied research, things on one-year-out time horizons.
After five months, I agreed to come in. I originally committed to three years. Now that’s been renegotiated. I’m committed to CableLabs for the long haul.
VB: What are you doing?
It’s about transforming CableLabs, and at the same time—the cable industry did not have a multi-gigabit strategy at that time. We had Docsis 3.0, where the top speed is one gig. We kicked off what became Docsis 3.1, which is now 10 gigs download and one gig upload. That’s now being deployed in the market – Atlanta, Chicago, all the members have some form of Docsis 3.1 deployment underway now. Then we did full duplex, which is 10 gigs symmetrical. That’s in specification work right now. It’s on track.
We’re trying to build CableLabs from near term to more long term. Today, the way we think about CableLabs, 50 percent of our technical spend is in one to three years out, and 50 percent is in three to eight years out. That three to eight years was almost zero dollars four years ago. It’s about CableLabs taking that much longer view, to be the source of those intellectual properties and innovations that drive the cable industry.
We do work in wireless access networks, like high-bit fiber coax, the traditional cable plant, fiber deep, fiber technologies. A lot of people don’t think of us as doing fundamental core research on everything from lasers to fiber materials and so on. We do a lot of work in security. We design and maintain the security infrastructure of the cable plant. That infrastructure is now being used by other industries. We provide those capabilities to open ADR, the automated demand response. These are the models going into heating and air-conditioning units. Utilities can tell you not to turn on your air conditioner, because it’s a brownout day. That’s an ADR module with an exclusive security key. We also do that now for health care and a variety of other industries.
We also have a broad influence on the industry standards organizations. We currently sit on 32 standards bodies, everything from IEEE to Open Daylight to OCF, you name it. In those cases, it’s about our contribution back to the broader industries around those technologies. Either the cable industry is directly interested, or they could impact the cable industry, like IOT devices that aren’t secure. All of a sudden they’re punching a hole in the network.
We also have a lot of work in AR and VR. We’ve been doing that for almost three and a half years now. We’re not inventing the AR and VR space, but we’re making sure that the networks those are going to run on are equipped and architected in such a way to take advantage and deliver great, compelling experiences. High speed, low latency tends to be a big area for us.
VB: How many people are in the R&D area?
McKinney: CableLabs today, total staff, I think it’s roughly 200 people worldwide. When I took over we were about a dozen members in about 18 countries. Today we’re 57 members in 33 countries. We now have some of our largest members in China. We have Japan, Singapore, other parts of Asia. Pretty much all of central and western Europe, a big chunk of eastern Europe. The U.S., Canada. We have 75 percent of Mexico and reach into Latin and South America. We’ve extended our international footprint.
Total technical resources, out of the 200, if you count all lab services into that number, you’re talking about 125, 130 total technical resources. A good chunk of those are all deep specialists with master’s degrees or PhDs. We don’t do the traditional “have a position, find someone to match it.” We just go find the best people in areas of interest. Then we create teams around them.
Take Peter Smyth, who’s now what we call VP of core innovation. We brought him in because he headed up all of wireless research for BT Research in the U.K. We have some of the top people in the world in their areas of expertise. We’ve spent the last four years building what I’d put on par with any of the top research labs out there.
VB: How do you look at things that would be considered competitive threats to cable?
McKinney: In many cases they may not be competitive. Everybody automatically associates Docsis with CableLabs. It was invented at CableLabs. It’s what gives you high speed over coax. And so people assume we must hate fiber. It’s actually not competitive, because we do a lot of work in fiber. We’re not pitting one technology against another. It’s about us deriving capabilities so that our members, the people who fund us – we’re funded by the capable operators – to give them the choices, based on their strategy, that will allow them to be the provider of choice in the markets they serve. Those markets can be high-speed data, wireless services, and so on.
Wireless is becoming one of the fastest-growing and biggest research areas for us. If you look at our 57 members in 33 countries, 24 of those cable operators also operate mobile networks. That’s Rogers and Shaw in Canada, GCI in Alaska, Liberty Global in Europe. Vodafone acquired two of our members, so now Vodafone sits on our board. Our involvement is really not about coax, coax, coax. All of our members have different strategies. Altice just announced that they’re going heavy fiber. We have great fiber updates. Comcast, Charter, they’re focused on Docsis 3.1 and eventually full duplex Docsis.
The only thing I’d probably say we’re not interested in is DSL. [laughs] Because of the copper pairs and the challenges there, that’s not an architecture that has a play with our members.
VB: What about 5G in general?
McKinney: Again, almost half our members are in the wireless space. Of those, they all have a 5G strategy. 5G has extreme low latency and extreme high speed. It’ll open up areas of innovation we can’t imagine. Mobile World Congress last year, we were there. At CES, you look at remote surgeries, remote control of robotics where you need that low-latency loopback through the network. All those things are great capabilities.
The thought that you’re going to deploy 5G in neighborhoods to be the broadband bypass—the challenge, one, is cost. How much do you pay for your 300 gigs a month on your cable bill versus 400 gigs on your cellular bill? Second is the challenge of cell strategies. 5G will be more of a small cell strategy. If you deploy those in neighborhoods, you have to look at the economics. You have to either acquire network assets somewhere in the asset base, local network assets, or you’re going to be doing a lot of trenching to go deep and serve those capabilities.
In our case, we have cable operators who are mobile network operators. You could argue – I don’t have any inside knowledge – why did Vodafone buy Kabel Deutschland in Germany? If you’re going to deploy 5G in Germany, you need network assets. Why is Liberty continuing to push on their wireless deployment? Both Charter and Comcast in the U.S. have announced that they’ve activated their relationship with Verizon. You’ll see some interplay with them between cable having coverage to 93 percent of all U.S. homes and where those 5G cell towers will be deployed.
VB: What about other directions, like over the top?
McKinney: Netflix, those kinds of guys? Again, a lot of people think OTT is a competitor. In reality it’s not. If you’re going to have OTT services, you’re probably going to step up your data package. If you look at where the biggest growth is in the cable industry, it’s around the growth of subscribers buying data services. You still see a slow decline of people dropping their video packages, but you see growth in data services.
Comcast announced their deal with Netflix. Now, if you’re on a Comcast X1, you have Netflix right in your set-top box. You can do integrated search using the voice remote, and when you pick a movie, it’ll ask whether you want to watch on a premium cable channel or on your Netflix account. It’s all the things consumers want. You’re going to see other cable operators do integrated work with Netflix and Hulu and all of those. It’ll become a bucket of everything the consumer wants.
At CableLabs we hold two conferences a year, summer and winter. Winter conference last year, a friend of mine is a senior executive at Netflix. He came out and gave the whole Netflix, what they’re working on, their streaming strategies, all that. We have 400 or 500 members in the audience asking questions, understanding Netflix, and he asked questions of us.
Silicon Valley has the typical “frenemy” thing going on, right? You’re a friend today and an enemy tomorrow, depending on the issues. But everybody has this perception about the cable industry and who their competitors are. I came in from the outside and had that same perception. People think of cable as being ESPN, CNN, Turner, whatever. It’s not. MSOs are the distributors. Cable programmers and the programmers. Distributors have to buy their content, and then they get the right to distribute.
When people complain about their cable bill—the average cost of content is increasing 8-12 percent per year. What Comcast pays for that content from ESPN or ABC or whatever, that goes up 8-12 percent per year. But on average, at best, they can pass on maybe four percent of that cost to consumers. That’s what consumers will accept. They still gripe and complain, but four percent, you can get that through. Who eats the difference? It’s the Comcasts and Charters of the world. They get squeezed. The cost of goods goes up, but they don’t have the price elasticity to pass that on. You’re seeing the margins on video products shrinking. But everyone likes to blame Comcast or Cox when their cable bill goes up.
VB: How does this wind up being a good business in the long run?
McKinney: It throws off a lot of cash. You look at any of the big cable guys, the cable stock, it’s stable. It’s predictable. Not a lot of downside risk. You’re not seeing anybody hitting a cliff. The cable industry is positioned very well competitively. It’s the leading broadband provider. Broadband becomes the thing that everybody wants access to, whether it’s in the home or even outside the home. Everybody wants access to Facebook or Twitter or whatever.
The ability for the cable industry to continue to invest in its network and maintain its leadership position in broadband services—you think about what the cable industry has invested over the last 10 years into networks, it’s hundreds of billions of dollars. It’s the largest privately-funded real estate project in the history of the United States, the cable network. Zero dollars from the government. From that perspective, the barrier to entry from somebody else to come into that market is high. Google tried. It was harder than they thought. These guys have been doing it since the ‘60s, or some of them back in the ‘40s, stringing coax.
VB: I asked Gary Shapiro what he thought about the Trump administration and net neutrality. His answer was, “It’ll come up, but I’m a little more interested in seeing other problems solved, like why the U.S. has high broadband costs compared to other countries, why there isn’t more competition.”
McKinney: Well, I think there’s a misunderstanding on pricing. We always see quotes asking why U.S. broadband speed is so slow compared to Japan or wherever. Here’s the thing you have to be careful with. The U.S. is viewed as one country, the same as Japan or Korea. If you take the U.S. and treat each state as a country and look at average broadband speeds by state, eight of your top 10 “countries” are U.S. states. Try to do high-speed broadband in Montana, Idaho, it’s tough. You have more cows than residents. When you dig deeper, we’re actually in a good position.
Now, with full duplex in the near future, 10 gigs symmetrical and no need to retrench anybody’s yard, that’s transformative. It’s not limited to the U.S. I was in China just three weeks ago, meeting with our members. They’re very focused on gigabit services. Every country in the world is focused on delivering the most efficient methods. There was a big study put out in the EU about the fastest way to get to gigabit services for residents in the EU. Docsis 3.1 with full duplex became the number one option in this study, because it uses the network that’s already invested. It requires zero construction. Change the cable modem, upgrade the CMTS and the other end, and you have 10 gigs symmetrical.
Everybody’s on the path. It’s this perception of one country against another country—what we did at CableLabs, when I first got here, there were three versions of Docsis based on geography. There was North America, Europe, and C-Docsis in China. The vendor community had to build three different products for three different parts of the world. When we rolled out 3.1, we got every region of the world, for the very first time, to agree on one version of that product. A vendor can build one product and ship it anywhere in the world. It’ll work on any cable network. That starts driving down cost, which is eventually passed to consumers.
When you can get this kind of global scale brought to delivering the technology solutions—I can sell an HP laptop anywhere in the world. 40 million laptops a year. I get that volume benefit and you see prices drop. That’s what we’re trying to bring to the cable industry. How do I lower those costs so I can get broadband services to a whole different level of demographics in the world today? A kid’s not going to be competitive on a global basis without access to broadband. We have to bring that cost down.
VentureBeatVentureBeat's mission is to be a digital town square for technical decision-makers to gain knowledge about transformative technology and transact. Our site delivers essential information on data technologies and strategies to guide you as you lead your organizations. We invite you to become a member of our community, to access:
- up-to-date information on the subjects of interest to you
- our newsletters
- gated thought-leader content and discounted access to our prized events, such as Transform 2021: Learn More
- networking features, and more