Martin Casado has been an icon of software-defined networking (SDN) — first as a founder of Nicira, then as the subject of Nicira's $1.26 billion acqusition by VMware, and then as a VMware fellow and CTO.

Now he's a senior vice president and the general manager of VMware's network and security business unit, a position announced just before VMworld. The job comes with the responsibilities of shepherding VMware's NSX into its next phase: sales and deployment. And he's really jazzed about that.

We caught up with Casado at VMworld last week to find out how his new job differs from the CTO role and to hear how SDN has changed now that it's a commercial reality. Read on to find out why he's excited about working with sales teams and why he thinks SDN is a "battle of titans" that goes beyond networking.

NSX is shipping for real and has a $100 million run rate. That must feel pretty good for you.

Casado: I feel like for the last seven years, I've been like, "Next year is the year for network virtualization! Follow me!" It's like I'm the Pied Piper: "One more hill and we'll get there!" And then the dam broke in the last six months. It totally, totally broke. A $100 million run rate in the last six months, 150 customers.

As general manager, you're replacing Steve Mullaney, Nicira's old CEO. Why did he leave?

Casado: The way he described it to me was, "I'm going in and looking at Steve Career 2.0." He's probably going to take some time off to figure out what that's going to be. He's been doing the day-to-day startup thing for 30 years, and he's just ready to move on. He'll probably be an investor. You'll notice he's on the board of Metaswitch.

So, you'll have to hire a CTO to replace yourself.

Casado: For now, I'm doing both. I'm a CTO and GM. It's interesting; most people view me as a CTO, because it's the capacity that —

Yeah, even right now I'm thinking of you as the "CTO" guy.

Casado: It's true! But if I wake up in the morning and think: What problems am I solving today? Almost all of them are business problems. That's where we are in the business. That's why I'm so excited about having the GM spot.

That would have been the case as CTO, too, right? You were doing a lot of customer meetings.

Casado: Yes, but the thing as CTO is: You're basically an independent contributor when it comes to sales. You're a guy that swoops in and has the happy conversation, and hopefully you're fun over dinner — but you're an independent contributor. You're not responsible for a number. It's not like you're managing a team. It's not like you're setting sales strategy or sales direction as much.

So now, I'm working with product marketing, I'm working with sales guys, I'm talking sales strategy, I'm talking how to enable the core [sales force] — it's just this business-level discussion. Which you may be overly bored with, but for me it's fascinating.

There's no question this [NSX] is becoming mainstream. So, this is the next phase. How do we turn this into a $1 billion business?

So, what exactly will you be doing as GM?

Casado: One of the reasons I'm excited about becoming GM is, I want to see a technology from initial inception to mass adoption. So I've been from the crazy idea phase, to the refined idea phase, to the iterate-on-the-idea phase, to the initial prototype and development phase, to the joint development and customer phase, to the product development phase, to the sales phase — I've done all parts of that.

So now, where are we? What are the primary challenges now for network virtualization adoption? It's just enabling sales forces.

I've got four different types of sales forces. The most advanced is me and maybe five other guys who have seen all this stuff and been involved in it — this set of guys, they can do an architectural sale. They can do a nuanced sale. They can get into every aspect of it — security and so on. For that, you can have big deals and get customers adopted. High success rate.

They totally don't scale. I did 320,000 miles on an airplane last year. I have no personal life. I've done 220,000 this year. You can't scale that.

The second sales force that I have is a set of specialists. We've got about 100 of those. These guys can go in and have an architectural discussion. But we're now approaching [having] a lot of customers, so even these guys are saturated.

The next sales force is the core VMware sales force. VMware's been selling to a mature market. Now, mature-market sales guys, they're tall, they're good looking, they have boats — you know what I'm saying? They're not used to selling on technology. They're used to the customer already being educated. They can talk value, but they need very specific use cases and specific ROI.

Microsegmentation is a major use case of ours, but one of the reasons that we've got it so refined is that we've got the challenge of enabling a core sales force.

The fourth sales force is the channel. In order to scale something to a $1 billion business, you need the channel.

So when I wake up in the morning and think, "What is the next stage of this journey," the next phase is actually enabling the sales force. It's just a cool place to be, right? Now we're not market-limited at all. We're not customer-limited. It's about sales capacity. Now that I'm GM, I want to focus on ramping that up.

How has the SDN market changed?

Casado: The broader SDN noise has died down. Before, I would go to Japan, and I would have to argue why overlay is better than hop-by-hop OpenFlow. We never have those discussions any more. I think the architecture war is pretty much over. Every legitimate network virtualization solution is overlay-based, whether it's Cisco, whether it's Nuage, whether it's us. I think people understand the architecture.

Cisco has gotten particularly aggressive with ACI this year. How has that changed things for you at VMware?

Casado: I think that if you want to understand the competitive landscape, the dialogue has been upleveled from networking to IT. There's basically three models of IT being pitched.

[First,] hardware-centric design. Which would be Cisco. They're not shy about this; they're saying, "Listen, you can accelerate things by ASICs. So we're going to give you something with a high-level interface that's also bound to a set of atoms" — a physical box.

Then you've got the software-centric pitch, which would be something like a Microsoft or a VMware or Red Hat saying, "You know what? Hardware should be general-purpose. And then you can consume stuff via software."

And then you've got customers that say, "You know what? To hell with both hardware and software — it should be a service. Your network of the future should be an Ethernet jack to the cloud."

The competition is at that level of discourse, and the reason is, we're all incumbents. So, now we're battling at the IT level, which is kind of exciting. This not about networking specifically; this has become a battle of titans.