The V.C. Corner: Shirish Sathaye, Khosla Ventures
So what's around the next corner in business tech? What's promising -- and what's overhyped?
In this occasional series, we sit down with the folks on the front lines to see what they're seeing.
Today, we talk to Khosla Ventures partner Shirish Sathaye.
ZDNet: What's going on right now?
SS: There are a couple of things going on. There's a lot of buzz around consumer-oriented companies and IPOs -- Facebook, LinkedIn, et cetera -- but there are four things driving a resurgence.
The first: more and more people are using the Internet much, much more than they did in the past. Enterprises are using it in very interesting new ways. What's happening is there's a tremendous shortage of bandwidth, inside and outside the enterprise. That's one trend.
The second trend is now, because virtualization has about 55 percent [penetration] in the enterprise, that is making people rethink the storage, network and computer architecture [around it]. The hypervisor is the key platform on which everything is going to be built. There are a whole bunch of companies ramping up in a way we haven't seen since the '90s. That's very exciting.
The third trend, a little overhyped in my opinion, is big data. The good news is that everything is generating data. The bad news is that you have to store that data and figure out how it can help your business.
We are sort of playing in all three areas. There are companies being created across the board.
The fourth trend is an increased focus on security. The new-age security companies are doing really well right now.
These are the four trends. You'll see young companies across all four. The incumbents as usual are sort of looking back, not ahead.
ZDNet: Let's talk about the first trend, Internet usage.
SS: A lot of the build-out in bandwidth is in the carrier network, the wireless carrier network. With the upsurge of BYOD, everybody has seven devices consuming bandwidth. That's causing tremendous strains in wireless infrastructure. The carriers are turning to Wi-Fi, LTE and 4G infrastructure buildups, causing a whole new set of companies to come out and take advantage of that. They sell to enterprises, but also to carriers.
It's absolutely an inevitable trend. People are using bandwidth when they're not even paying attention. That's causing problems at the edge of the network, where people need to get content, but also inside the carrier's network. And they're starting to use enterprise equipment out of necessity, rather than some desire to embrace Wi-Fi.
ZDNet: Which brings us to your second trend, around supporting infrastructure.
SS: Here's one example. Way back when, EMC, Network Applicance and Sun put storage on one end of the network, networking on the other end of the server, and put the disks in the middle? And that was great.
They write things sequentially. It was fast. But when you're running a performance environment, [it puts tremendous pressure on storage performance.]
We have a company in this space called Nutanix that's doing extremely well. This is completely new thinking. The hypervisor is turning everything on its head.
ZDNet: You say big data is a trend, but you do so holding your nose a bit. Why?
SS: Big data is an important trend, but people are using the term big data in ways that don't apply. It's very important because we have all of these things now that create gobs of data.
If you watched 60 Minutes two or three weeks ago, they ran a story on the F-22 Raptor, which was grounded by the Air Force for a couple of months because, inexplicably, pilots feel dizzy when they fly it. There is so much data in the flight patterns and the computers; there should be a way to analyze the data to figure out what's going on. There's plenty of oxygen in the cockpit.
How do you solve this problem? You don't know what questions to ask. You need to ask specific questions. You need a whole new set of technologies to look at this data in a holistic way. And then there's using it to understand customers better, buying patterns better, like the Target-knows-you're-pregnant story. We just don't have that technology. The old database method just doesn't work.
It's real. It needs to be developed. There are companies doing that. One company, Palantir, is being used a lot by the three-letter agencies to analyze data. How do you know when there are two bad guys talking [on the phone], and not Andrew and Shirish? That kind of inference is useful in finding patterns that are unusual.
ZDNet: And so your fourth trend is security.
SS: It has become an issue of national security and enterprises. There's a whole set of hackers all over the world, in many cases sponsored by governments, that are coming up with new kinds of attacks that go through firewalls and go through many of the intrusion detection and prevention devices out there -- it's called an advanced persistent threat, using one computer to attack others, causing significant chaos.
As more and more things get networked -- the grid, nuclear facilities, but even enterprises -- they won't even know that something's going on.
I saw a demo of an RSA conference where a guy showed how he could walk around and impact insulin meters and kill patients without ever going into the room. More and more things are getting computerized. You need a whole new set of companies to address this.
ZDNet: How is the economic climate affecting all this?
SS: For the young companies, the present economy doesn't affect them. When these companies mature, who knows what the situation will be in one or two years.
I'm worried about two things: one is Europe; who knows what's going on over there. Greece, Spain, France -- I have no idea. The quicker they resolve that issue, the less damaging it will be.
The other issue that bothers me is the extremely high valuations in the consumer space, being paid for companies. We've been there before, in '98 and '99. And then it turns around, and people overreact on the downside. It's a gut feeling. At some point, the valuations will have to correct. And they always overcorrect. I worry about the effects of the impact on that affecting enterprise companies.
It's not isolated. After the U.S, Europe is the biggest market for enterprise right now.
ZDNet: Are there any other concerns you have that aren't related to the economy?
SS: VMWare has 85 percent market share in the hypervisor market share, which is great, but 85 to 90 percent market share generally doesn't last very long -- well, I take it back, it could last 7, 8, 9 years -- but it's already lasted that long. So that's something I feel a little bit uneasy about. It could be 2, 4 years from now, but I just know that some things don't last.
There are many companies banking on that and developing to the VMWare platform. These companies would have to do other things.
ZDNet: And then there are the cross-currents between consumer and enterprise.
SS: What used to happen in the old days, if I wanted to start a project in the enterprise, I'd have to fill out a form and hand it to IT and they'd give me a server after three or four months. Now they've started building private clouds that have quickened the development of things in the enterprise.
ZDNet: What keeps you up at night?
SS: It's Europe, and how that will affect the world financial markets. That [situation] has an impact on exits of more mature companies looking at the IPO market, and buyers who may not have the currency to acquire my companies.
Just get Europe fixed. Please.