SM: You were at Zilog, but their focus was not on networking. EB: Right. We attempted to build a business out of the networking developments. When I say we, I say a few friends who all ended up being significant contributors to the networking industry. They are people like Judy Estrin, Joe Kennedy, Bill Carico, people who ended up starting all of these great companies in the late 1970’s and early 1980’s.
We attempted to build a commercial networking business at Zilog, but we were not given the green light, which discouraged the entrepreneurs amongst us. In 1979, Charlie Bass, who was my immediate boss, left Zilog and founded Ungermann-Bass with Ralph Ungermann. This was the first Silicon Valley networking startup. I was supposed to join him except I had started my green card process and I was told by my lawyer I should not change employment in the middle of the process, otherwise it would have to go back to the start.
Ungermann-Bass got started, raised money, and unfortunately, I could not join them. It turned out to be a good thing, because when my green card came through about 18 months later, I was able to start a company called Bridge Communications which ended up being the second networking company competing directly with Ungermann-Bass in the same general space. Our two companies were the networking pioneers.
SM: What was the industry landscape like then? Was it easy to raise venture capital? EB: There was a little bit, but it was not easy to find. The 1970’s were not particularly dynamic for the economy in general. It was after the oil crisis shock of the late 1970’s, but there were some excitement around computers and microprocessors. Of course the Internet was not known. We ended up raising money from Lawrence Weiss Peck and Phil Greer, and you know the history of that firm. Now it is Light Speed Ventures. At the time we raised $1.8 million.
SM: Lot of money for the time. EB: It was an incredible amount of money for me. I had no idea you could raise this kind of money. It was enough to allow us to start our company, build our team and develop our first product. I remember we got the go ahead in the fall of 1981, and we closed a few days before Christmas of that year. We got started right away, and 1982 was a big development year. Christmas of 1982, exactly one year later, I was personally installing a first beta customer, UCLA. Basically UCLA had wanted to network all of its computers, which were PDP 11 and VAXs.
They were considering Ungermann-Bass but we got them excited about a new product that was substantially superior architecturally and performance wise. The only condition was it absolutely had to work by the time the students got back after the Christmas break. That meant we had to fix all the key bugs in about 2 weeks while the students were gone. I spent that Christmas break there, and it was a very, very close call. The network was not bug free, but it worked well enough that when the students got back in January they had access to not one computer, but the entire network.
They would type something like “connect me to VAX” and the network would go out there and find the first VAX available with sufficient bandwidth. In those days, that was magic. Obviously it enabled the departments to dramatically improve the utilization of its computer infrastructure. It became a reference site and we soon had many other installs. We bootstrapped the company in the engineering, scientific and University community as they were the early adopters of networking technology.
We had a very sound architecture which enabled us to continue building on existing products. We had open APIs which enabled us to leverage pieces coming from third parties. We built a substantial product line in a short time. This is still before TCP/IP and the Internet. At the time, we had not bet on TCP/IP because it was some kind of University protocol, and was not a terribly good one.
The best protocols to run networks in those days were the ones built at Xerox called the XNS protocols. They had a better architecture, they were open, and the specs had been published.
However, by 1984 TCP/IP had become sufficiently popular, in part because it came with UNIX, that although it was deemed the inferior protocol, people were asking for it. We had to create a new version of our software with TCP/IP. It was a good thing we did that because it enabled us to penetrate the market right at that moment. We took Bridge public in 1985. It was a very successful IPO. The company at the time had about $35M in revenues.
SM: Enormous for the time!