Shulou(Shulou.com)11/24 Report--

On April 7, 1946, the year after the end of World War II, a baby boy was born to an ordinary family in Brooklyn, New York.

The baby boy's father is a technician in a gyroscope factory. And his mother is an ordinary housewife.

For this family, just after the twists and turns of the war, to be able to have a son, is a very happy thing.

The couple had high hopes for the child, hoping that when he grew up, he would rise to the top as a top engineer (engineer was a very important profession at that time).

Sure enough, the baby boy, Metcalfe, lived up to their expectations. When he grew up, he showed great talent in learning. In 1964, he graduated from high school and was successfully admitted to the top American school (and the cradle of top engineers at that time), the Massachusetts Institute of Technology.

Metcalfe's parents would not have thought that their children would not only become a top engineer, but also set up a Fortune 500 company. One of his great inventions not only changed the lives of countless people, but also had a far-reaching impact on the trend of the IT industry.

Yes, this Metcalfe is the inventor of Ethernet technology, the founder of the famous technology company 3Com, and the creator of Metcalfe's law, Robert Melankton Metcalfe (Robert Melancton Metcalfe).

Robert Melankton Metcalfe █ rookie was rejected for his doctoral thesis in 1969. The 23-year-old Metcalfe graduated from the Massachusetts Institute of Technology with degrees in electrical engineering and business administration. A year later, he received a master's degree in computer science from Harvard University and went on to pursue his doctorate at Harvard.

While studying for his doctorate, Metcalfe got a job with the MAC program at MIT. The MAC project team, which specializes in operating systems, computing theory and artificial intelligence, became very famous.

In 1969, the US Department of Defense promoted the construction of ARPANET (the predecessor of the Internet), which connected the mainframe computers of four famous schools.

The University of California, Los Angeles, the University of California, Santa Barbara, Stanford University and the University of Utah Metcalfe followed the incident and thought it was very interesting. So he strongly advised Harvard University and MIT to connect their mainframe computer systems to ARPANET. Metcalfe is both a graduate student at Harvard and a researcher at MIT. )

The arrogant Harvard University rejected his proposal, but MIT agreed.

Soon, Metcalfe completed the construction of the network interface, connecting the MIT mainframe computer to the ARPANET.

Based on his design and research, which established a high-speed network interface and protocol software between the ARPANET IMP and PDP-10 time-sharing minicomputers, he wrote a doctoral thesis and submitted it to the Harvard University's degree committee.

In June 1972, Metcalfe failed to defend his doctoral thesis because the degree committee considered his thesis to be "mathematical" and "theoretical".

The blow is not just from Harvard. While participating in the ARPANET project, Metcalfe led 10 AT&T (AT & T) officials to visit the ARPANET demonstration. As a result, the system crashed during the demonstration.

Metcalfe wrote in his recollection:

"I looked up painfully and saw that they were laughing at the unreliability of packet switching. I will never forget that. for them, this confirms that circuit switching (the technology used in traditional fixed-line phones) will continue to exist, while packet switching is an unreliable toy that will never have much impact on the business world."

Metcalf was a little frustrated by one blow after another. Soon after, he received a warm invitation from Bob Taylor (Bob Taylor, one of the main sponsors of Arpanet), director of the laboratory at Xerox's Palo Alto Research Center (Palo Alto Research Center), to join the lab and finish his paper. Metcalfe gladly agreed.

The Palo Alto Research Center is the famous PARC laboratory.

PARC Lab gave birth to many great inventions, such as laser printer, mouse, graphical user interface (GUI), bitmap graphics and so on. Many of the innovations of Jobs' Apple computer come from here.

After █ was inspired by his peers and came to PARC Lab successfully by the project, Metcalf soon started his work.

At the time, PARC Labs wanted to design the world's first personal computer (later known as Alto). Metcalfe's task is to design a network interface for the computer so that they can be connected to each other.

To build a multi-user terminal computer network, the biggest problem is how to coordinate the access and occupation of each computer host to the network.

In the early 1960s, computer scientist Leonard Cranrock (Leonard Kleinrock) proposed that the mathematical queuing theory could be used to coordinate the data flow in the network by simulating traffic jams and people queuing.

ARPANET adopted this theory and proved it to be effective.

In 1971, Norman Abramson (Norm Abramson), a professor at the University of Hawaii, set up a radio network called ALOHAnet (ALOHA is a common greeting for Hawaiians), using a more "radical" scheme than ARPANET.

In ALOHAnet, data is transmitted in the form of tiny packets. It does not attempt to avoid collisions between packets. On the contrary, any user who loses a message due to a conflict only needs to retry sending a packet after a random interval.

How ALOHA works. It's like two people talking. If both sides start talking at the same time, both sides will stop immediately. Restart the conversation later. After several attempts, there will always be a situation in which one party does not speak, and the problem is solved.

You will find that there is an obvious flaw in ALOHAnet's strategy: it is a waste of resources, which is useful in low-traffic situations, but when the network becomes congested, conflicts become more frequent and transmission efficiency decreases significantly.

After reading Norman Abramson's paper, Metcalf was deeply inspired. Soon, he improved ALOHAnet's model and proposed a new model.

In the new model, the computer host independently adjusts the waiting time of transmission retry based on the collision frequency. If the number of conflicts is small, they will quickly retry; if the network is congested, they will exit to maintain the overall efficiency of the communication.

Metcalfe's new model makes up for the shortcomings of his paper. Soon, in May 1973, he finally passed the Harvard reply and received his doctorate. (it's worth mentioning that Metcalfe didn't publish his paper at Harvard, but at the Massachusetts Institute of Technology. )

In his research project, Metcalfe also introduced a new model.

On May 22, 1973, Metcalfe issued a memo called Alto Ethernet, formally proposing the idea of Ethernet (Ethernet).

In the memo, Metcalfe sketched how Ethernet works. "participating sites, such as AlohaNet or ARPAnet, inject their packets, and they travel at megabits per second, with collisions, retransmissions and backwards," he suggested. "

Part of the memo page Metcalfe's Ethernet vision combines Norman Abramson's random retransmission mechanism, his own adjustments to the system clock and other improvements to the ALOHAnet model to mitigate the impact of conflicts.

Some of these theoretical innovations were developed by other researchers, but Metcalfe was the first to integrate them into actual network design.

It is necessary for Xiao Zaojun to explain the origin of the name Ethernet.

Before electromagnetic waves were discovered, ether was proposed as a ubiquitous transmission medium (light is transmitted through the ether). Later, it was discovered that the ether did not actually exist.

Metcalfe adopted the name "Ethernet" and regarded Ethernet as a medium. He himself got a nickname, "Ether Daddy".

In June 1973, Metcalf was allowed to build a 100-node prototype Ethernet.

To complete complex tasks such as logic design, building circuit boards and writing microcode, Metcalfe enlisted David R. Boggs, a graduate student at Stanford University, to help.

Metcalfe (left) and Boggs (right), thanks to their efforts, the world's first Ethernet prototype system was officially born on November 11, 1973.

At the time, the Ethernet transmission rate reached 2.94 megabits per second, about 10,000 times faster than the previous terminal network.

After █ founded 3Com to promote the popularity of Ethernet technology, Metcalfe strongly suggested that Xerox could commercialize the technology. However, the response of the management of Xerox is very slow and there is no real action.

In 1979, Metcalfe, who had been waiting for six years, couldn't stand it and left the PARC lab. He decided to start his own company to promote the popularity of Ethernet technology. The company he founded was the later famous communication network enterprise-3Com.

The name of 3Com company comes from three letters: computer (computer), communication (communication) and compatibility (compatibility). This fully reflects Metcalfe's desire to improve the compatibility of computer communications.

After the establishment of 3Com, the commercial feasibility of Ethernet was greatly improved by selling network software, Ethernet transceivers, and Ethernet cards for minicomputers and workstations.

In 1980, brokered by Metcalfe, DEC, the world's second-largest computer company, Intel, a semiconductor company, and Xerox, a large system provider, formed a technology alliance to launch the DIX (the letter of the three companies) Ethernet standard.

Soon after, in 1983, IEEE set up a special working group to launch the IEEE 802.3 standard based on a variant of the DIX standard.

The earliest 802.3, 10BASE5, had only the throughput of 10Mbps, the medium was thick coaxial cable, and used the access control method of carrier sense multiple access with collision detection (CSMA / CD, which science and engineering students should be familiar with).

In addition to Ethernet, IBM and General Motors have also launched their own network standards.

In particular, the token ring technology of IBM has formed a fierce competition with Ethernet. Finally, after 20 years of competition, Ethernet won and the token ring was eliminated.

After the success of █, the transformation of personal retirement, Metcalfe promoted the popularity of Ethernet throughout the 1980s. On March 21, 1984, 3Com was successfully listed.

In the mid-1980s, Metcalfe put forward an important point of view, that is, "the value of a network is proportional to the square of the number of nodes of the network." This point of view is the famous "Metcalfe Law".

This law has important reference value for understanding the network effect and the development of Internet economy.

Metcalfe left 3Com in 1990 to become a critic and technology columnist.

Metcalfe of that period also made a lot of mistakes. Some of his strange remarks often make himself the object of ridicule by the media.

In 1995, Metcalfe thought the Internet would suffer a "catastrophic collapse" the following year. He said that if his prediction was wrong, he would "eat" his words.

Later, it turned out that he was indeed wrong. In 1997, at the sixth International World wide Web Conference, he took out a printed copy of his article and put it in a blender with some liquid. Then, amid the cheers of the audience, he calmly ate the stirred paste. While eating, he admitted his mistake.

Metcalfe's prediction errors also include: Linux will be killed by Windows2000; wireless networks will be abandoned by the mid-1990s; and in 2006, Windows and Linux will not be able to handle video services.

The unreliable prophet, Metcalfe left the media in 2001. He founded Polaris Venture Capital and turned into a venture capitalist. In 2011, Metcalfe went to the University of Texas at Austin as a professor.

In 2022, more than 50 years later, he returned to the Massachusetts Institute of Technology's computer Science and artificial Intelligence Lab (formerly MAC, now renamed CSAIL) as a researcher.

During this period, 3Com has also experienced a lot of storms.

In 1999, 3Com's revenue peaked at $5.7 billion. But soon, the dotcom bubble burst, 3Com fell off the altar, and the market capitalization shrunk sharply. In November 2009, 3Com was acquired by Hewlett-Packard for $2.7 billion in cash.

In his later years, Metcalfe won numerous awards for his great contribution to Ethernet.

In 1996, he was awarded the IEEE Medal of Honor. In 2003, he received the National Technical Medal and the Marconi Award. In 2007, he was inducted into the National Inventors Hall of Fame.

Not long ago, on March 22, 2023, Metcalfe, 76, was awarded the 2022 Turing Award by the American computer Association (ACM) with a prize of $1 million (from Google).

Jeff Dean (Jeff Dean), senior vice president of research and artificial intelligence at Google, said in an official statement from ACM:

"Today, there are about 7 billion network ports around the world. Ethernet is everywhere, and we are used to it. However, it is easy to forget that without the invention and efforts of Bob Metcalfe (that is, every computer must be connected to the Internet), our connected world would not be what it is today."

█ concluded Jeff Dean was right. Ethernet technology is the cornerstone of people's data communication network. As the father of Ethernet, Metcalfe's contribution is enormous.

Today, Ethernet is still the main standard for global wired network communication. Its processing data rate has been upgraded all the way from 2.94Mbps and 10Mbps to 400Gbps, 800Gbps, and even 1.6Tbps.

The familiar figure of 802.3 will stay with us for a long time until it is replaced.

All right, that's all for today's article. Thank you for your patience.

This article comes from the official account of Wechat: fresh Jujube classroom (ID:xzclasscom), author: Xiaozaojun

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.