Q&A: Silicon Valley Ubermensch Andreas Bechtolsheim explains what the big deal is with cloud computing


This article was originally on a blog post platform and may be missing photos, graphics or links. See About archive blog posts.

Andreas Bechtolsheim, co-founder of Sun Microsystems and chief development officer of Arista Networks, a Silicon Valley cloud computing company. Credit: Alex Pham / Los Angeles Times.

Google made waves in the tech world this week when it announced plans to release an operating system that would encourage wider use of something called cloud computing.


Although most have never heard of cloud computing, many do it every day. By uploading photos to Facebook, sending messages via Gmail or playing Club Penguin online, users are accessing programs and software files that live far away in cavernous, climate-controlled rooms containing thousands of computers.

To help explain this shift in the way we use computers, we turned to Andreas Bechtolsheim, co-founder of Sun Microsystems and chief development officer for Arista Networks, a Silicon Valley startup that supplies networking equipment used to build these massive arrays of cloud computers.

As it turns out, Bechtolsheim was also one of the first people to invest in Google back in 1998, when the company was just two Stanford geeks with a laptop. His $100,000 investment in the company started by Sergey Brin and Larry Page, along with several other shrewd calls, turned the Birkenstock-wearing engineer into a billionaire.

We spoke to the 53-year-old serial entrepreneur recently about cloud computing, his investment philosophy and his latest venture, Arista Networks. An edited version of the conversation is below.

Q: What do you make of the potential for cloud computing, both as a market and a technology?

Bechtolsheim: It is a surprising evolution in the history of computing. Every application can now shift to the Web. You can access any application remotely. My startup does the networking plumbing for this.
IDC has estimated that by 2012 the market for cloud computing infrastructure will grow to $42 billion, up from $16 billion in 2008. It’s the fastest-growing slice of the spending on information technology. Right now, it’s small sliver of the overall pie. Most of the spending is for the applications. But it’s a growing slice of the pie.


Q: What are some uses of cloud computing?

Bechtolsheim: Hollywood uses high-performance clusters to ...

... render movies. The advantage of the cloud is that it’s always running and not idle. They get much better utilization. These compute farms can simulate car crashes, render movies. Facebook also does an enormous amount of computing going on behind the scenes to generate relevant content to you. That’s what is going on inside the cloud farms. What you see is just what is relevant to you.

Q: As a consumer, what’s the benefit?

Bechtolsheim: You get convenience. You don’t have to deal with servers, data storage or big expensive computers. The laptop accesses everything. You don’t have to worry about backup or security; it’s all being handled by the company that manages the cloud.

Q: Who are some of the key players in this space?

Bechtolsheim: The leader is Amazon. They started in the business by renting out their idle servers during non-holiday-shopping seasons. Google provides applications in the cloud. Facebook is a cloud company; they host a number of third-party applications on their cloud.

Q: How did your initial investment in Google come about?

Bechtolsheim: This was in 1998, and I was using the Internet to search for things. Alta Vista was the search engine then, and they had a very simple algorithm. It counted the number of times words appeared in the document. So people built these dark pages behind the Web page to game the system. As a result, the search results became useless, and doing a high-quality search was impossible. I was just interested in getting better search results. After I heard from Larry and Sergey about their idea, I rushed out to my car to get my checkbook. The company didn’t exist yet. So I wrote them a check and said, “Here’s a check to get you started.”

Q: You’ve made a number of smart calls over the years. What do you look for in an investment?

Bechtolsheim: When you invest in a company, you want to see a viable business plan where it can be profitable over time. In the late 1990s, there were a lot of companies that basically just took an idea for an existing business and put it on the Internet. They took dog food, for example, and sold it online. That wasn’t a very good business model. Whereas Amazon created something completely new that you couldn’t find a single physical bookstore.

Q: What else?

Bechtolsheim: Simplicity. If an idea makes intuitive sense in the first five minutes, it’s probably a good idea. But the more time it takes to understand, the worse it gets. Google was a very simple idea.

Q: Tell us about your latest venture.

Bechtolsheim: Arista Networks addresses a void in the market for simpler, faster and more scalable Ethernet switches. Cisco dominates this market, but their products have become increasingly complex and expensive. Our focus has been on creating software that can be used to build large networks that play well together and are very reliable. In essence, we are building better plumbing that is more scalable and more redundant [than competing switches]. We do it by solving a fairly basic problem that hasn’t been solved in 10 years.

Q: As a non-engineer, I almost hesitate to ask what that problem is.

Bechtolsheim: People are building cloud systems that have tens of thousands of computers. It sounds simple, but the key thing here is cost and how you hook it up so computers can all talk to each other in a way that minimizes latency. But the existing network protocols don’t scale well. So we came up with a way that does. For every challenge, there is a solution.

-- Alex Pham

Follow my random thoughts on games, gear and technology on Twitter @AlexPham.