Interview with Gordon Bell
An important aspect of Ethernet’s beginnings is that it was not simply a clever idea—it also was a necessary one.
After forty years since the birth of Ethernet, Gordon feels so strongly about Ethernet to claim it has become a fundamental component to almost EVERY computer: Processor, Memory/Storage, Internet Port, Operating System, and Application hardware/software.
Gordon Bell had been with Digital Equipment Corporation (DEC) in the early 1960s where his achievements included major contributions to architecting the company’s Programmed Data Processor (PDP) line of minicomputers (though they were explicitly not called “computers” because of that term’s association at the time with prohibitive cost and complexity). He left in 1966 to join the computer-science faculty at Carnegie Mellon University and then, in 1972, returned to DEC as vice president of engineering. In this role, Bell oversaw development of DEC’s historic Virtual Address eXtension (VAX) computers.
The VAX line would prove hugely popular in the scientific research communities and influential in bringing about the computer age across varied industries which had sought a less expensive and more flexible and nimble computing capability than previously available. But a key problem had to be solved: how to connect the devices.
“We had two or three different networks that were in place within our research group but none that was exactly like Ethernet.”Gordon Bell
“In the fall of 1978 we had predicated the ‘VAX Strategy,’ that stated we would seamlessly intermont all of the computers to form a single VAX/VMS compatible programming environment…clusters of DEC’s highest performance computers, departmental minis, and personal workstations,” Bell said.
“…This hierarchy of computers was all connected by a line we called the network interconnect, but we had no idea exactly what that interconnect was going to be. The diagram that I sold the VAX Strategy on essentially looked like it had a single bus that connected all of these computers and through the levels.”
“We were desperate since the whole plan was based on ‘an Ethernet,’ and we didn’t have it. It was clearly a bet your company plan,” Bell said.
DEC’s critical need led to Bell’s conversations with Ethernet’s father, Robert M. Metcalfe of Xerox’s Palo Alto Research Center (PARC) in January 1979. Those conversations led to a virtual handshake between DEC (in Boston) and Intel (in San Francisco) over AT&T’s first Picturephone Meeting Service (PM) and the formation of “DIX.” And that eventually led to the launch of IEEE Project 802, and the IEEE 802.3 Standard and DIX’s first draft ( Blue Book) in September 1980—and the decades of evolution that have unfurled across the sprawling Ethernet landscape and which still continues to today.
“We saw having a standard as being very important,” Bell said. “The alternative was… we would have had to ‘wait’ and deal with all of the slaw evolution and idiosyncrasies of the communications industry. We would’ve been in communications-protocol hell, and that was not a place you wanted to be.”