Join top executives in San Francisco on July 11-12, to hear how leaders are integrating and optimizing AI investments for success. Learn More
The high-tech and venture capital (VC) communities continue to generate charismatic leaders, but few can compete with Bob Metcalfe, co-inventor of Ethernet at Xerox Parc and cofounder of local-area networking pioneer 3Com in the 1970s.
While still an actively developed technology, Ethernet is overshadowed now by closely-related Internet and totally-unrelated Ethercoin technologies. But it seeded a new world of connectivity.
[Follow along with VB’s ongoing Nvidia GTC 2022 coverage »]
Metcalfe’s spirited efforts to push forward high-tech and VC innovations still bear fruit. Today, the Association for Computing Machinery (ACM) named Metcalfe as recipient for the 2022 ACM A.M. Turing Award for the invention, standardization and commercialization of Ethernet.
Join us in San Francisco on July 11-12, where top executives will share how they have integrated and optimized AI investments for success and avoided common pitfalls.
Metcalfe walked with no little swagger from his grade school days, when he told a teacher he would go to MIT — which he did — to his days at fabled Xerox Parc, where he named Ethernet after the imagined substance Newton used to describe a transmission medium for the propagation of electromagnetic forces. Metcalfe showed fierceness and flair in the LAN battles that pitted 3Com against the likes of IBM, Wang, Ungermann-Bass, Interlan and many others.
Metcalf followed his time at 3Com with forays into publishing — he was CEO, publisher and pundit for InfoWorld Magazine — and VC community building in Silicon Valley, Boston and Austin. These days he is an emeritus professor at The University of Texas at Austin and a researcher at MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL).
He spoke with VentureBeat just ahead of formally accepting the Turing Award. (Editor’s note: This interview has been edited for length and clarity.)
VentureBeat: Your work along David Boggs on Ethernet took some cues from Norm Abramson’s ALOHANet, forged some performance enhancements and met some skepticism there at Xerox Parc. But it really now looks like a classic case of how things seem so obvious afterwards. People asked: ‘Why can’t I connect two computers in the same room, if I can connect them over vast distances?’
Metcalfe: Clearly, the difficulty to connect two computers in the same room was an opportunity. But our network’s first competitor was ‘SneakerNet’. People said ‘Why should I spend $1,000?’ And that usually was its initial cost. They’d say ‘I just carry a diskette over to the printer and print it out.’ So, we had to fight SneakerNet for a while. Eventually the competition shifted to other methods of networking.
Ethernet arose in 1973 when Xerox Parc decided to make what is — arguably — the first modern personal computer. People thought it was outrageous. They said: ‘You’re going to put a computer on every desk — why would you do something so stupid?’ Fortunately, I got the job of networking them together. We were building a laser printer that ran at a page per second, 500 dots per inch, which meant it required lots of bandwidth. So, to keep the printer busy, we needed a fast network. And the first Ethernet was 10,000 times faster than what it replaced, which was terminal networks like RS-232. So, we could keep the printer busy.
VB: Looking back, Ethernet seems like a case where network standards trumped proprietary standards.
Metcalfe: IBM [Token-Ring] and General Motors [MAP Token bus] and Wang [WangNet] and others all decided that they wanted to dominate networking with their own technology. We had a big fight that lasted maybe 20 years. And we used the IEEE to standardize our technologies. And three of them got standardized: Ethernet, IBM Token Ring and Token bus. But Ethernet won that battle. And my company prospered as a result of standards.
There is an irony in the fact that IBM and Wang and others were contending. That meant that the PC makers were reluctant to choose between them. So, instead of putting the network on the motherboard, which they should have done very early, they didn’t. And that left open an opportunity for my company to sell network interface cards that would plug into those slots and give them networking capability. Pretty soon we were shipping those cards by the millions. The price went down and the volume went up. And our company grew into the billions. Then we and the industry evolved. Along came TCP/IP, and protocol differentiation went away. We all adopted TCP/IP, and got ourselves on the Internet.
VB: What George Gilder dubbed ‘Metcalfe’s law’ has become very influential. It dealt with a network’s value growing as device numbers increased. Now, such network effects are under scrutiny as social media grows. How do you view the effect of computing and networking on societies?
Metcalfe: I think networking has over-delivered. In a short 50 years we’ve reached three quarters of the human race and we’re doing so with ever-rapid increase — so much so that connectivity has overwhelmed us. We don’t know what to do with it.
A number of pathologies have developed — you may remember the first pathology of the Internet was pornography. And they had to pass an act of Congress — the Communications Decency Act — to deal with it. And then along came advertising, which for a while was viewed as a pathology. But then we realized it was going to finance the entire Internet. And then came spam, which was a pathology, and we’ve pretty much handled spam — almost. Then, we have fake news.
My view is that we have a series of pathologies that we handle as they arise. But the real cause of our problems is that we don’t really know how to manage connectivity quite yet.
VentureBeat’s mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Discover our Briefings.