Custom Search
Entrepreneurial Capitalism and Innovation:
A History of Computer Communications 1968-1988
By James Pelkey

Entrepreneurial Capitalism & Innovation:
A
History of Computer Communications
1968 -1988
By James Pelkey

This history is organized by three co-evolving market sectors and also standards making.
An overview of the schema is presented in the Introduction.

DATA COMMUNICATON
Ch. 1: Emergence
Ch. 3: Competition
Ch. 5: Market Order
Ch. 11: Adaptation

NETWORKING
Ch. 2: Vision
Ch. 4: Arpanet
Ch. 6: Diffusion
Ch. 7: Emergence
Ch 8: Completion
Ch. 10: Market Order

STANDARDS
Ch. 9: Creation

INTERNETWORKING
Ch. 12: Emergence

 

 

Chapter 7
Networking: Emergence 1979-1981
LANs and DataPBXs

7.11     Robert Metcalfe and the Founding of 3Com

Metcalfe left the LACN Symposium more certain than ever that Ethernet had to be freed from the proprietary grasp of Xerox if it was to have a chance of impacting the coming commercialization of local area networking. Then came the welcomed news that the letter Gordon Bell of DEC had sent to three Xerox executives had generated interest. Metcalfe remembers:

"Finally I got a bilateral meeting to happen between Gordon Bell and David Liddle.

And in the midst of this, I went to the National Bureau of Standards on another consulting contract. I was told the day before my visit, a guy who I have never met had been through from Intel, and he said he had some 25 megahertz chip technology and was looking for applications. And as soon as I heard this I had an idea: 'Why don't we just build an Ethernet chip with it?' So I called up Intel and arranged a meeting.

When I arrived, I was introduced to a guy named Phil Kaufman, who it turns out reported at the time to Andy Grove. They had assembled thirty to forty people for me to tell them a little about Ethernet and what they ought to do with it. I would normally get paid for consulting but I did it for free that day because it would have been embarrassing for me to leave. So I pitched them on how they ought to develop a custom chip for Ethernet. And the reason they ought to do it is that if DEC and Xerox were going to do this, they're going to need chips and you guys can make the chips for their standard.”

Kaufman, already a believer that communication chips would be significant "consumers of silicon," remembers:

"I was looking around at communications and the issue, obviously, came down to what protocols should be used. Now, the one protocol that had actually been used somewhere was Ethernet, as done at PARC. Everything else was inventing on the fly. We took a good look at Ethernet and all the papers that had been written and said: 'It doesn't matter whether it's good or not, it exists and it works. Let's see if we can take off from there."

Liddle remembers being contacted by Kaufman:

"Intel contacted me and was interested in making a networking part primarily for IBM. They wanted to make a single part that could do both SDLC and Ethernet, and they wanted it to be variable speed and run real fast and all this kind of stuff. They were anxious that it should be at least ten megabits, maybe faster.

reasoning was very simple. Dr. Wang and IBM and all these people were after us, Xerox, because we had been running these very successful TV ads and were scaring everybody. Competitor claims were getting made that: 'Oh yeah, but it’s really costly and it always will be.' So I said: 'If I want to get stuff that's inexpensive, first of all I've got to show that I've got a semiconductor manufacturer who says: 'We'll make all the components so they'll be cheap' and secondly I need a down and dirty equipment vendor who makes modems and terminals and throw-away stuff, which was certainly Digital's reputation at that time.

So I thought this will get away from the image that this is a premium high-end expensive product, and it'll encourage it. I wanted to start an industry at that moment. I didn't want any proprietary network. I wanted to be able to say: 'Look, I can hook to all this other gear that other people make.” I didn't want to have a post hoc networking industry develop."

Xerox and DEC invited Intel to join their meetings. Kaufman remembers:

"What our attorneys essentially told us was that we could not write a three-way agreement to work on the project to develop the technical specification even though we were explicitly going to put it in the public domain, and that we could not have tri-lateral meetings if there were any marketing people present from any company."

Metcalfe soon learned of the impasse. Concerned that lawyers might end the collaboration; he called his former college roommate, Howard Charney, an anti-trust expert, who had just finished suing IBM successfully on behalf of Memorex. Armed with the new knowledge that: "All you have to do is have a meeting whose purpose is expansion of markets and making of standards to be publicly disclosed" to avoid anti-trust problems,” Metcalfe again intervened, facilitating a three way meeting. [6]

Freeing Ethernet from Xerox was but one of Metcalfe’s concerns that May of 1979. A second, and more emotional, issue was how to avoid being left behind in the coming commercialization of local area networking. As a consulting company of one, he knew he either had to grow a company and begin selling product or become marginalized. For others now believed as he did and they had every intention of starting companies to sell local area networking products. How did he know? Others were asking him to join them.

Yet Metcalfe knew his vision of networking desktop computers differed from the views of others - even as he also knew no existing desktop computers needed to be networked. Yet believing that if he did not stake claim to market leadership, the confusion created by others could leave him stranded, left behind and out of luck, an outcome he had no intention of letting happen.

On June 4th, Metcalfe and Gregory L. Shaw incorporated 3Com Corporation (3Com); for computers, communications and compatibility. Incorporation changed little. Metcalfe knew he needed a client to fund a growth in headcount, adding the pith and substance 3Com required to transition from consulting to a product company. Metcalfe remembers:

"I went to General Electric (GE) and said: 'You should do home networking.' They said: 'Well, all right. Sure. Let's work on it for a while.' Consulting to GE is probably the deal that made the company because I was able to expand from one person to ten people, and in particular I was able to add Howard Charney. [7] And eventually we got into designing a personal computer related to it. The personal computer thing got shelved, but we kept working on the home network called HomeNET. They paid us about three or four hundred thousand dollars over that time.”

In addition to his consulting, Metcalfe continued to shepherd the three-way cooperation among Xerox, DEC, and Intel to make Ethernet public. Finally, that fall of 1979, the three firms met at the Boxborough Sheraton, Boxborough, MA. Acknowledging Metcalfe’s essential roll, he came to the pre-meeting dinner. The following day the technical teams of the three companies met, without marketing people or Metcalfe.

Metcalfe’s cares now revolved around 3Com and how to fund products that would conform to the technical specifications he confidently expected DEC, Intel and Xerox (to become known more familiarly as DIX) would negotiate and then publish. Metcalfe remembers: [8]

"We went from GE to Exxon. Exxon, in a deal that, I don't know how I did it, but we got Exxon to agree to pay us to develop our first three products. We gave them fully paid up non-exclusive licenses and they allowed us to retain ownership of the technology. The contract was written that they could only use this technology in connection with an Exxon system. The three products were: TCP for UNIX, the Ethernet transceiver and the Ethernet adapter for the Unibus."

3Com had a contract to build the three essential components for an Ethernet network: an adapter to connect to the computer, a transceiver to interconnect to the media, and protocol software to enable usefulness. The market focus would be interconnecting DEC Unibus computers. [9] And while the UNIX operating system was not as widely used as DEC’s proprietary operating systems, UNIX was used often enough and porting TCP to UNIX had already been done by others. Why TCP for UNIX? Why not XNS, the protocol architecture developed under Metcalfe at SDD specifically for Ethernet? Metcalfe explains:

"In 1980, 3Com Corporation was about to do its first product, very first product, which was to be some protocol software. And, Vint Cerf persuaded me, it didn't take much persuading, but persuaded me, that we should implement TCP, because it was going to be a standard. Now, at that time XNS was a secret of Xerox, and I had just left Xerox. I couldn't do XNS, even if I wanted to, but TCP was public domain. It was to be a standard."

In May 1980, DEC, Intel and Xerox (i.e. DIX) made it official and announced they would publish their joint specification of Ethernet on September 30, 1980. Metcalfe knew he needed to raise significantly more capital than could be accumulated through profits if 3Com was to make the product and organizational investments required of a market leader. To raise the capital meant selling equity capital, a percentage of 3Com ownership, to venture capitalists (VCs). After writing a business plan, Metcalfe began meeting with VCs. Only venture capitalists refused to properly value the fact that 3Com was a going organization that had three "real" products.  Again Metcalfe:

"All the venture capitalists said: 'No, this technology doesn't matter because Exxon, you see, has a fully paid up non-exclusive license. They could just put you out of business.' And I said: 'You don't understand. They're not going to do anything with this technology. I mean, the chances that they're going to compete with us in this technology are pretty small. And by the way the contract was written they could only use this technology in connection with an Exxon system.' They said: 'Ah, no, no, no. You're first round.' And I said: 'No, no. We're second round, and we own all this technology.' 'No, no, you're first round.'"

The crux of the argument: How much ownership would have to be relinquished to raise capital? The venture capitalist argued for as low a company valuation as possible so their investment would buy the maximum ownership. After all, they reasoned, 3Com’s revenue for the fiscal year ending May 1980 totaled only $281,000 and the company had sold rights to its technology to Exxon. So how real a company was it?

On September 30, DIX published the Ethernet specification. Metcalfe felt backed into a corner and agreed to accept venture capital on the terms offered. But the venture capitalists made a new demand. They wanted to see an experienced executive hired to bring management competence to the largely engineering-staffed 3Com. Metcalfe, partly defensive and partly desperate to raise money, challenged the VCs to find the successful executive.

Gib Meyers, a general partner of the Mayfield Fund venture capital partnership, responded by asking Metcalfe to meet Bill Krause, a "rising star" at Hewlett Packard (HP); Krause had managed HP's successful minicomputer business from $50 million to $500 million in sales. Krause remembers:

"I had just gotten knee deep into an assignment of getting HP into the PC business and had developed a strategy for HP to use Ethernet technology to network the PCs to become a competing alternative architecture to the HP 3000, which I launched. In other words, 'better to have HP eat its own young rather than somebody else,' and so Gib called and said -- and he knew that I wanted to start a company and had had several conversations with him -- so he said "You ought to meet this guy that we've been talking to that's involved in starting a local area network company," and I said: "Fine.  Who is it?" And he said: "It's Bob Metcalfe." I said: "He's the inventor of Ethernet, you know, God, he's an industry leader. I'd love to meet him."

After a series of meetings, Krause made the leap and resigned from HP to join 3Com. He recalls his logic:

"I had not met somebody as charismatic and smart as Bob Metcalfe, with one exception, Howard Charney, and I had not seen as strong a technical group as Ron Crane and Greg Shaw. So the first reason, as is always in venture capital, was the people. The second reason was that Bob and I shared exactly the same vision of how Ethernet was going to eventually be used for PCs. Third is that I really did think it was going to be the third wave of computing and I always thought that if you got in front of the mainstream of something, a lot of the rest was made easy.  I had joined HP just as they were getting into the minicomputer business, so that just made me getting on the fast track at HP a lot easier, being in front of something, of the new wave."

With Krause joining as President, and Metcalfe assuming the CEO role, three venture capital funds, Mayfield, New Enterprise Associates and Melchor Venture Capital, invested $1.1 million on February 27, 1981. Paul Baran, who had joined 3Com’s Board of Directors in March 1980, invested as well. 3Com’s post-money valuation totaled $7.6 million, suggesting Metcalfe had prevailed in the battle over valuation. (Post-money valuation represents the total outstanding shares of a company’s stock times the price shares were sold at. Typically a first round financing would value the company at less than $5 million and generally less than $2 million. [10]


[6] The author has chosen to present Metcalfe’s version of history. Liddle, who is as equally credible, remembers the issues clearly as those of intellectual property, not anti-trust.

[7] Metcalfe also recruited Ron Crane,  who had worked on the ARPANET report with Baran and Cerf.

[8] The irony is that Exxon had local area network technology in its Zilog operations and yet turned to an outside, start-up company for its technology.

[9] A computer bus, or backplane, is the communication network interconnecting the different subsystems of a computer; the subsystems existing as printed circuit boards physically plug into slots interconnected by the bus. Examples of then buses were Unibus, Multibus, Q-bus.

[10] In calculating the number of shares outstanding, preferrred shares are treated as if converted to common shares.