Light Reading founder and CEO Steve Saunders talks with Colin Doherty, the CEO of data center interconnect specialist BTI Systems.

Steve Saunders, Founder, Light Reading

April 7, 2015

23 Min Read
CEO Chat With Colin Doherty, BTI Systems

Data Center Interconnect, or DCI, is one of the hottest sectors in telecom currently. Since coming back to Light Reading last year, prodigal-son style, I've consistently heard about one player in the DCI space punching above its weight: BTI Systems.

Recently I took the Acela, (a.k.a. "the world's slowest high-speed train") up to Boston to meet BTI Systems Inc. 's CEO, Colin Doherty, to hear more about what makes his company tick.

Colin is an industry veteran who joined BTI after selling Arbor Networks to Danaher Corporation in 2010. Previously, he served at Mangrove and Nortel.

Figure 1: Colin Doherty, CEO of BTI Systems Colin Doherty, CEO of BTI Systems

We covered a lot of ground in a short space of time -- from customers' preferred purchasing criteria for DCI platforms, to the overabundance of standards for virtualization (turns out you really can have too much of a good thing), to the coming shakeout in the metro optical market.

The interview took place at the Boston offices of one of BTI's chief investors, Bain Capital Ventures, and for bonus points Bain Partner Deepak Sindwani popped in to chat about some other interesting topics -- including the prospects for a telecom IPO, the nascent drone communications market and the absolutely critical issue of whether cellphone service on planes would be a good or a bad thing.

Enjoy.

Stephen Saunders, Light Reading: Hey Colin.

Colin Doherty, BTI Systems: Hi Steve.

SS: It was interesting for me coming back to Light Reading last year and meeting BTI again, because I met you guys before, in the early 2000s, when you sold optical components. But this time around everyone I met was new, and the entire purpose of your company seems to have changed. So what gives?

CD: Yeah, back in 2011 a number of people in the company had the right idea to try and differentiate the company from what I would term "the optical path," where everyone competes on speeds and feeds and bits and bytes, when really what the new market is looking for is network flexibility and usability. It's not all nailed-up circuits anymore. It's gotta be a little bit more flexible than that.

The new vision was to differentiate the company towards a software and application-ready platform. And to have that run on a flexible, agile, optical base. And it turns out that lines up nicely with the current industry consensus that you want a virtualized and simplified Layer 2, Layer 3 platform.

So you are running on this flexible, agile, optical, electronic layer with a layer of management orchestration on top. And that's the BTI strategy. And that's when they hired a lot of new people into the company, around '11, '12.

SS: At one point the company was selling components, right. When did the switch to systems happen?

CD: A lot earlier. I mean, I joined the company in April of '14. BTI had a good stable of Tier 2, Tier 3 carrier customers for its systems, mostly in North America, for a while before I got here.

And then they identified, early, the opportunity in the data center interconnect space: the cloud market and the web operators who started to build their own networks and not put traffic on the carrier's network. They recognized the need to get content closer to consumers' eyeballs, right? So the metro traffic began to grow to a point that in 2018, maybe 62% of total IP traffic's gonna be in the metro.

So BTI identified that market nice and early, in '11, and then built a platform to address those evolving needs of the new carrier.

SS: What's the platform called again?

CD: The 7800 Intelligent Cloud Connect.

SS: OK. And when you say "Layer 2, 3," the 3 in there is IP?

CD: Yeah. So you take some MPLS core capability and you flatten that and you put it on an optical platform. So this platform has a number of different things in it that are a requisite for intelligent cloud connect. Inside that platform you have transport, you have optics, you have a backplane and a control plane. You have management. You have a packet fabric.

SS: Do you have security?

CD: Not yet.

SS: But?

CD: But you have a packet fabric, you have software, and then you have application-ready blades, as opposed to just having transport, optics in a pizza box. That's the speeds and feeds point-to-point connectivity solution. That's fine for certain things. But if you wanna put a solution in and scale extensively well, our platform has 3.36 terabits of capacity.

But it also has software applications, a packet fabric, manageability, and virtualization. That's the direction that we see it going in.

So, to your security question, could you develop encryption software for low-latency financial markets? Yes. Could you port other application and develop applications on these open compute blades? Yes. And I think that that was the vision that the team had. And I think they're right. You see the emergence of a lot of point-to-point, high capacity speeds and feeds boxes.

SS: So many.

CD: That's fine. But again, networks aren't built anymore on point-to-point, nailed up circuits. It's about flexibility and usage and usability of that network. That's what the platform allows you to do.

SS: And it's proven really popular.

I hear about BTI wins all the time, wins which are being taken from some really big companies. So congratulations on that.

But I think there's probably some misconceptions around BTI and data center interconnect. And I want to dig on this with you a little. Because obviously when I think of BTI now, I think of data center interconnect, or DCI. That's what I associate your company with. Are you happy with that?

CD: You know, I think it's a positive aspect to what we do. I don't think it's the only thing we do. I think there's really three things that build a balanced business from point of view of markets. So data center interconnect, cloud providers, colocation guys, ISPs, web providers: they move quickly. They deploy big pipes. They want to run flexibly, intelligently. And they want to run the network hotter.

Their deployment's quick. They have rapid services innovation. And we're a small company, so we can move quickly with them. That's one market.

The second market that's important is the carrier market in general: the Tier 2, Tier 3 guys in North America. That's still a very good, growing market. They're taking some lines from the Tier 1s as the Tier 1s in North America focus more on how they compete against the web providers, potentially by accessing large enterprises and building competitive networks. So the Tier 2, 3 carriers for us are important. That's still a large part of our business.

The third leg of the stool, if you like, from a balanced revenue model, is the Tier 1 side.

And part of our challenge is that, as a smaller company, you have this domain approach in North America. We certainly have spent time on developing partnership and channel relationships to try and address the Tier 1s with a new platform.

SS: How long is the process to get into a Tier 1 account now? Because it used to be like four or five years.

CD: I think it's probably shortened a little bit. It's hard to say how much. We have partners, resellers, that we work with today that give us opportunities to get this platform into the Tier 1s.

SS: Are you allowed to say who they are.

CD: One is Fujitsu.

SS: Okay. Great company. Big customer of ours. Interesting.

CD: But again, data center interconnect -- absolutely we're happy to be associated with that. We lead the field. Some research earlier this year put us as No. 1 in DCI, which is nice. But the other markets are also areas as focus for us.

SS: Is DCI getting you to a really healthy place financially and funding the opportunity in the other markets?

CD: The pace of change in the data center market is what attracts a lot of people to us. You're beginning to see people that used to be long-haul providers now turn to the metro space. So there's a lot of competition and focus, a lot of publicity and press and media around DCI.

Today, if you look at one of these new phones, the latest software update on one of them is a 5 gig download. Long-haul networks used to be run on 10 gig, 40 gig, and then jump to 100 gig. So you take a metro area and everybody's downloading the new software, that 10-gig pipe's not gonna work. Right?

That pace of change and growth is a fit for us, because we're relatively small, nimble. We can support people. We can respond quickly.

CD: It's part of the reason. Not the entire reason. But we can ship quickly. We're not a 16-week lead time company. The data center web cloud colo environment deploys faster. They make decisions quicker. And a lot of the services work that we do is to build these networks for these players. So there's a service component here as well as a technology component and a platform component.

I think the pace and the deployment and the pace of change, the innovation that these companies demand, they all fit our model. We're a small, nimble company.

SS: What are the key purchasing criteria for DCI equipment?

CD: You need more than just point-to-point capability. Data centers are not just point-to-point environments anymore. You may have six to seven to eight data centers in a metro. You may need to change traffic loads and transfers between those points multiple times.

You may need to change bandwidth deployments at different times of the day, different peak times. You may need content in different locations. You need the sophisticated orchestration of management and virtualized traffic feeds to be able to do that.

Not to say that it's not valuable to have the lowest cost per bit on a point-to-point circuit, but I just don't happen to think that we are going to win that game. And I don't think that's a high margin game either. So when you get to services, packet, and software and applications, I think that's where you get potential larger margin and a broader profitability play for the company.

SS: DCI is obviously not the only hot part of optical or telecom at the moment. There's long-haul, 100-gig transport SDN, for instance.

CD: We're not a long-haul company.

SS: Not interested?

CD: No. BTI has always been a metro regional player, which is why we have a rich set of Ethernet services. We have a 10-gig wave product called the 7000 that's been in the market for some time. We have a very flexible, usable, open management system that allows some of these content guys to develop their own interfaces. So long-haul is not a market we're gonna go after at all.

SS: It's good to be focused, isn't it?

Let's talk about SDN and NFV. The only thing getting more attention than DCI is the concept of virtualization. Are SDN and NFV real in carrier networks right now?

And when do you expect to see those things really take hold?

CD: Yeah, good question.

SS: Well, that's why they pay me the tiny dollars.

CD: I think that they're real. Are they deployed en masse today? I don't think so. But I think there are real conversations happening inside carriers about: "How do we change our infrastructure? How do we flatten it? How do we become more nimble? How do we use a virtualized infrastructure with Layer 4 through 7 on top of that to orchestrate faster deployment of services and higher margin?" So I do think those conversations are happening in a much more concrete way than they had been over maybe the past four or five years.

We think that there are green shoots of optimism there on change, in conversations that we see and have with players in the market. So yeah: I think that there's a need for the new DCI web cloud colo guys to start attacking other parts of the carrier customer base. And I think the traditional carriers are gonna have to respond to that.

SS: And virtualization is one of the ways they can do that.

CD: Yeah. You can't build networks the way they used to be built. And I think that there's now a real ecosystem of meaningfully capable technology around to change things.

SS: I thought it was telling at Mobile World Congress a couple weeks ago when Deutsche Telekom made their big announcement of publicly available NFV-based services in Croatia. So it's like fantastic for the Croatians. I'm sure they're delighted.

A lot of people assume, looking at the media coverage and the analyst coverage, that virtualization is everywhere right now. And of course it's not.

And one of the things which I'm concerned about having dug in a little bit here is the ever-increasing number of standards organizations, quasi and de facto and official, involved in developing specifications, not just for virtualization, NFV and SDN, but also sort of the upskilling of IP and its integration with OSS and BSS and management and MIMO and all of this sort of stuff. There's like 40 or 50 different industry groups all working on it. And, to me, having lived through the development of many, many different networking standards that looks spectacularly ill advised.

Here we're trying to do something really hard with the technology, incredibly complicated. But everybody wants to have a whack at it. Everybody wants to have a committee. Talking to Deutsche Telekom they said they are cutting back on the number of standards organizations they are taking part in because they can't possibly keep up.

I suspect that there's an interoperability train wreck coming. Because you're going to have all of this different work going on. Nobody can keep up with all of it. And the only reason we haven't seen the first examples of a lack of interoperability is that it's not on the network yet.

Do you think that I'm wrong to be concerned about this? What is your take on the whole standards miasma at the moment?

CD: No, not wrong. And it's an interesting question.

I think that the pace of change in life -- industry, social media, open compute storage, traffic growth, content, security, all of those things -- mean we need a lighter standards infrastructure than was utilized in the past.

We have decided to build the platform on, if you like, third-party building blocks, right? We didn't put $50 million into an ASIC development.

So I think the opportunity we have for standards differentiation is to create as open a platform as possible to allow those players that want this virtualized, simplified Layer 2 and Layer 3. And then you can write and provide applications on top of that.

So relatively open.

SS: So you're saying the openness needs to be customer facing. Not as much between vendors. That's a significant point you're making.

CD: Yeah. [Vendors] are building networks that have all proprietary protocols and systems and standards. And then if you want to move quickly, you can't, because you're handcuffed to that protocol or that structure or that infrastructure you built. And again, credit to the BTI team that changed this in '11. I think the vision and architecture was correct: that a lot of the new operators are fundamentally content application software companies. They're gonna write their own code on top of a flexible platform that allows them to scale applications, software, services, over packet and multiple transport networks.

It's incumbent upon us to provide that platform that is as open as possible to allow people to port and code on top.

SS: When I was interviewing John Chambers recently he said he thought that in the future his biggest competitors would be white box solutions, as opposed to Alcatel-Lucent, Juniper, Ericsson, and so on and so forth. Is that a trend that you see emerging?

CD: In the content web environment, you are going to see the request for the white box virtualized high-speed capacity platform where these players can write software on top.

SS: But in your metro world, you give them that ability but you also give them all of this other stuff.

CD: Yeah. I think, in our world, we need to be able to provide a lot of the packet and the software and the services as well. Because that is margin-enrichment. As opposed to a bare steel white box that's more about high speed compute blades.

SS: Or a simple switch function.

CD: Right. But I do think that that is a trend that's going to emerge. And I think the complexity of the existing carrier base that provides local line rural telephony and all of the existing infrastructure that people kinda poo poo a little bit, but... SS: We don't poo poo it. Lots of money there.

CD: Huge. And it's a great service.

SS: So true, they come to our shows, park their tractors, take their boots off, sit down, and it turns out that they've got these really vibrant networks which have leapfrogged incumbent infrastructure in many places. We have a huge amount of time for them.

CD: The existing carriers have done a terrific job in trying to evolve from all of the regulatory things that they have to comply to. But it's tougher for them to move as fast in this open compute environment.

SS: There's a lot of anxiety there. We see that a lot.

Let's talk about the metro market. To me, it just looks like there's got to be a shakeup there. There's an awful lot of equipment manufacturers, obviously.

CD: Yeah. I think that there are an increasing amount of players in that space. And again, we think the ability to differentiate, for us, it's not speeds and feeds.

We think it's around intelligence, flexibility, and usability of the network. But, of course, if there's nobody in a market, you're in a bad market.

SS: But how many do you think there are right now when you go in for an RFP? 20?

CD: For metro, I wouldn't say it's as big as that.

SS: 12?

CD: I think for optical there may be 20. But for metro, it may be closer to six to eight maybe. Something like that.

SS: So Ciena, Infinera, or Cisco, Alcatel-Lucent, all the usual suspects?

CD: Uh-huh.

SS: It must be really annoying if you're an incumbent, right? Because you've got this business and it's been around forever. And then you guys come along. My business, Light Reading doesn't have very many competitors and the ones we have are all going out of business. So that's convenient. But if I had a small, really good hot competitor that came into my market and was winning business from me, I'd find that very irksome. Do you sense the irk coming from your competitors?

CD: I think that they have turned their attention much more to the metro space than they did in the past. You asked the question before about long-haul: Do I see that as a place we're going to go? No. We're absolutely not gonna be a long-haul company. Some players are in submarine, some in ultra-long-haul, some are in long-haul. We've always been in metro. But I think they've turned their attention to the fact that the building block inside metro now is 100 gig.

That used to be the bit rate of choice for long-haul. Now it's standard building block in metro. So I think it's a natural place for them to evolve to.

But it's not really about putting a chassis in place and putting a controller and a line card and then waiting for that to fill up over time as the long-haul traffic slowly fills. It's about deploying internationally, globally, very quickly, fully populated products with a lot of flexibility and virtualized management -- where traffic is constantly changing. So I think you need a rich Ethernet set of services.

Again, you've got Ethernet services being driven by PDAs and phones and downloads and people watching movies. You sit on a rental bus in an airport at any point? Everybody's listening to something, watching something, reading something, sending something.

And these are all driving Ethernet services that carry Ethernet up towards multiple 10-gig e-ports. And then the content above that is at 100-gig increments. So it's natural for the long-haul guys to come into metro. But the service set that you need is different.

SS: So let's say there are eight metro players for the sake of argument -- serious players. Is that four too many?

CD: It's certainly a lot. It'd be great if there was only two or three.

And I think that people are going to choose different products for different reasons. You're not gonna win them all. But we've probably won our unfair share against really large players.

SS: Which makes it an exciting story.

[There's a knock on the conference room door, and Deepak Sindwani, Partner at Bain Capital, comes in].

Deepak Sindwani: Just wanted to say "Hi."

SS: Hi. Steve Saunders. [Steve and Deepak shake on it.]

DS: Hey. Deepak Sindwani. How are you?

SS: Yeah, good. Come in.

DS: Mind if I join for a few minutes?

SS: Yeah. Please do.

DS: I don't wanna be recorded.

CD: [Points to two voice recorders on table] You are being recorded.

DS: Just kidding. I don't care.

SS: How long has Bain been involved with BTI?

DS: We've been involved since 2011. We invested in March of '11.

SS: [Points to Colin.] So you must be happy with this guy?

DS: I've known Colin a long time. Many lives, actually.

SS:< Are you doing other investments in telecom?

DS: Yeah. We look at communications software more than systems these days.

SS: That's the trend.

DS: If [we are] going to be in networking, it would be in the software-defined networking area. And that's where these guys [BTI] are headed. Other solutions like marketing and SAAS solutions that are sold into telecom, that's a big space that we're spending time on.

In the network equipment space, I have personal investments in companies in the mobile infrastructure area, because that world is flipping over significantly.

SS: Yeah, the 5G stuff is gonna change everything, isn't it?

DS: In the hard core systems space, obviously we love where BTI's situated because of the metro cloud. That's exploding. I don't think we'd invest in a basestation company or like a 100-gig long-haul optical business. It's just a tough market to break into as a young company. But we pick places where we think innovation can disrupt. Just like these guys are doing with intelligent cloud connect.

BTI identified the whole metro cloud space early on, and identified that [the customer could be] Internet companies, or the hosting businesses. Clearly, the biggest guys are gonna bring it in-house. The other guys are gonna use Rackspace or Equinix or somebody who's also one of our customers. So they said "let's just draft off of that, build a product that's purpose-built for them." And we know that the guys who lag the most in that space are the Tier 1s. But they're eventually going to get there.

And so I give them a lot of credit for seeing that trend early on and working backwards and saying what's the disruptive architecture?

And I think CD and team's approach of label switch routing and optical and then application layer capabilities makes sense. Customers don't need to have a million-dollar Juniper or Cisco router sitting next to something from Infinera that is kind of a pure optical box.

And so it's a big bet. But that's what venture's about. It's about making a bet on a company like this that can change the network infrastructure.

SS: So what's your end goal for BTI?

DS: CD's gonna make this a big successful company.

SS: Do you think telecom systems companies can IPO still?

DS: Recent history would suggest that Cyan's had a hard time. But do I think the market would be receptive to the metro SDN company? Yeah. I do.

SS: Quick question: Have you tracked drone communications at all?

CD: I have not.

SS: There's lots of drone comms companies. DHL and Amazon and Facebook and Google are throwing billions at that market right now. It just really interesting for me.

DS: Yeah. I don't get it. I think they're all gonna build them. And I think they're gonna work, although some will probably crash. And then I think the [Federal Authorities] are going to see "We're not ready for this yet."

SS: Well, the word coming out of the FAA is that they will allow high-altitude drones because they don't want all the investment to go overseas. But we'll see. I mean, it's cool stuff. I like anything which is grandiose and disruptive with elements of science fiction.

DS: I just don't [agree with] Jeff Bezos' view that my Amazon delivery's gonna be delivered by a drone by, like, 2020. I don't see that. Imagine some drone flying down the street in the South End.

CD: Well, there wouldn't be one. There'd be like hundreds of thousands of them.

DS: Like mosquitos.

SS: Well, I think the future is in the heavy iron, the big bombers that are up at 65,000 feet and stay there.

DS: I totally agree with that.

SS: Semi-satellite orb, gigabit speeds. We have the technology in the lab now to put 400 pounds of mobile wireless equipment in a drone which can provide gigabit capacity down for backhaul application. That's relevant.

DS: Absolutely agree with that.

SS: But that's 15 years away.

DS: I still don't see those mosquitos zipping around.

SS: No. I don't either. Those are toys.

DS: It's like having cellphone access on airplanes. That would be a disaster. Can you imagine?

CD: That is going to happen soon.

DS: Internet access is useful. But having everyone on the phone next to you, on a flight? Listening to somebody talk on the phone.

SS: Dreadful.

DS: They used to have those air phones, but no one used to use them because they were like $3 a minute.

CD: If not more.

SS: Yeah, and they didn't really work. So they were good for hitting a person in the face next to you when they're on their cellphone.

DS: Yeah. I would just reiterate: we're big fans of CD and of BTI. And we had our four year anniversary. So the company has more than doubled since we've been involved. And it's not a bad thing.

SS: Cool. Well, I think we're due to go for some lunch. Your timing was fantastic.

CD: [Points at Steve Saunders's leather-bound journal, which was a gift from BTI's marketing department.] I like the book.

SS: This, I'm very happy with. This gets good reviews. I'm a connoisseur of free leather-bound journals. And this one has traveled the world with me for the last three months. It was in Barcelona a couple of weeks ago. I had it in front of John Chambers. You could tell he wanted one.

DS: Let me tell you something: you're special, because I've never even gotten one of those.

CD: Really?

DS: No BTI leather-bound book for me.

CD: The chosen few, Deepak. The chosen few.

About the Author(s)

Steve Saunders

Founder, Light Reading

Steve Saunders is the Founder of Light Reading.

He was previously the Managing Director of UBM DeusM, an integrated marketing services division of UBM, which has successfully launched 45 online communities in less than three years.

DeusM communities are based on Saunders' vision for a structured system of community publishing, one which creates unprecedented engagement among highly qualified business users. Based on the success of the first dozen UBM DeusM communities, the UBM Tech division in 2013 made the decision to move its online business to the UBM DeusM community platform – including 20 year old flagship brands such as Information Week and EE Times.

Saunders' next mission for UBM is the development of UBM's Integrated Community Business Model (ICBM), a publishing system designed to take advantage of, and build upon, UBM's competitive strengths as a leading provider of live events around the globe. The model is designed to extend the ability of UBM's events to generate revenue 365 days of the year by contextually integrating content from community and event sites, and directories, to drive bigger audiences to all three platforms, and thereby create additional value for customers. In turn, these amplified audiences will allow business leaders to grow both revenues and profits through higher directory fees and online sponsorship. The ICBM concept is currently being discussed with a broad group of business leaders across UBM, and is earmarked to be piloted in the second half of 2013 and early 2014.

UBM DeusM is Saunders' fifth successful start-up. In 2008, he founded Internet Evolution (www.internetevolution.com), a ground-breaking, award-winning, global online community dedicated to investigating the future of the Internet, now in its fifth year.

Prior to Internet Evolution, Saunders was the founder and CEO of Light Reading (www.lightreading.com), Heavy Reading (www.heavyreading.com

Subscribe and receive the latest news from the industry.
Join 62,000+ members. Yes it's completely free.

You May Also Like