Vint Cerf

Vint Cerf: The Grammar of a Connected World

The internet is not a network. It is an agreement between networks that were never designed to speak to each other, and the terms of that agreement were largely written by two men in 1973. Their 1974 paper described a host-level transmission control system and gateways that would route between dissimilar networks — an architecture so durable that it still carries essentially all global traffic five decades later. One of the authors, Vint Cerf, has spent the rest of his life in a three-piece suit explaining it to the world.

Van Nuys High School

Vinton Gray Cerf was born on 23 June 1943 in New Haven, Connecticut, to Muriel Gray and Vinton Thruston Cerf. His mother's family was Canadian, of British, Irish and French Canadian descent; his father's had emigrated from Alsace-Lorraine to Kentucky. He is hard of hearing, and has been throughout his life — a fact that shaped both his personal history and his professional advocacy. He met his wife Sigrid at a hearing aid practice in the 1960s; they have two sons, David and Bennett. He has served on the board of Gallaudet University, the university for deaf and hard-of-hearing students, since 1997.

Van Nuys High School in California produced an improbable concentration of talent: Cerf's classmates included Steve Crocker and Jon Postel, both of whom became central figures in the creation of the internet. While still a student he spent six months at Rocketdyne working on the Apollo programme, writing statistical analysis software for tests of the F-1 rocket engine. He took a bachelor's degree in mathematics from Stanford in 1965, worked two years at IBM as a systems engineer supporting QUIKTRAN, and returned to study at UCLA, taking a master's in 1970 and a doctorate in 1972 under Gerald Estrin.

The Room Where ARPANET Started

At UCLA he joined Leonard Kleinrock's data packet networking group — the laboratory that housed the first node of the ARPANET. Cerf worked on the host-to-host protocols for those earliest connections. It was hands-on, foundational work at exactly the right moment: the network existed, it was small, and the rules for using it were still being invented by the people using it.

In 1972 he became an assistant professor at Stanford, where he researched the problem that would define his career: how to interconnect packet networks that had incompatible designs. Radio networks, satellite networks and wired networks all had different packet sizes, different error characteristics, different assumptions. Beginning in 1973 he worked with Robert E. Kahn on an architecture to make that irrelevant.

The solution was elegantly indirect. Rather than forcing every network to conform, Cerf and Kahn proposed a common protocol layered above them, with gateways at the boundaries translating between one network and the next. Their 1974 paper set out the design; in December 1974, with Yogen Dalal and Carl Sunshine, Cerf published RFC 675, the first detailed specification of the Transmission Control Protocol. By 1978, working with Jon Postel and Danny Cohen, the design was refined into the arrangement still used: a simple Internet Protocol responsible for getting packets from one place to another, and a more complex Transmission Control Protocol above it responsible for making that reliable. TCP/IP.

From 1973 to 1982 Cerf was at DARPA, where he funded the development of TCP/IP itself, packet radio networking, packet satellite networking and security technologies — controlling the money as well as writing the specification.

Making It a Public Thing

Cerf's second career was in making the network real for people outside research. At MCI, where he was vice-president for digital information services from 1982 to 1986 and senior vice-president for technology strategy from 1994, he led the development of MCI Mail, which in 1989 became the first commercial email service connected to the internet — a decision that helped break the network out of academia.

In 1992 he co-founded the Internet Society with Kahn and served as its first president, and later as chairman of its board. He helped establish ICANN, the body that administers the internet's domain names, joined its board in 1999 and chaired it from November 2000 to 2007. Since October 2005 he has been vice-president and Chief Internet Evangelist at Google, working chiefly on global internet policy and standardisation. He was president of the ACM from 2012 to 2014, and a presidential appointee to the National Science Board in the same year.

His long-range project is the Interplanetary Internet, developed with NASA's Jet Propulsion Laboratory: a delay-tolerant networking protocol designed for a medium where signals take minutes or hours and links break constantly. It was installed on the International Space Station in June 2016. Since 2015 he has also warned about digital obsolescence and the prospect of a digital dark age, in which the documents of our era become unreadable because the software that interprets them has vanished.

The honours are essentially complete: the National Medal of Technology in 1997 with Kahn, the IEEE Alexander Graham Bell Medal the same year, the Marconi Prize in 1998, the Turing Award in 2004, the Presidential Medal of Freedom in 2005, the Japan Prize in 2008, the Queen Elizabeth Prize for Engineering in 2013, foreign membership of the Royal Society in 2016, the IEEE Medal of Honor in 2023 and the California Hall of Fame in 2024, along with honorary doctorates from more than forty institutions.

Why Vint Is Called a Genius

The Turing citation names two things, and the pairing is the point: "pioneering work on internetworking, including the design and implementation of the Internet's basic communications protocols, TCP/IP, and for inspired leadership in networking." The design insight was the recognition that the hard problem was heterogeneity, and that heterogeneity should be solved by adding a layer rather than by imposing uniformity. TCP/IP works because it assumes almost nothing about the networks beneath it — a discipline of deliberate ignorance that let the internet absorb technologies, from dial-up to fibre to mobile radio, that did not exist when it was written.

The second half of the citation, inspired leadership, is unusual language for the ACM and reflects something real. Cerf has spent fifty years as a diplomat for an architecture — at DARPA distributing funding, at MCI making it commercial, at the Internet Society and ICANN building governance for a system with no owner, at Google arguing policy.

The counter-case is straightforward. The work is shared with Kahn, and every major award reflects that; Postel, Cohen, Dalal, Sunshine, Crocker and Kleinrock all contributed materially. The internet was a well-funded government programme, not a lone insight, and packet switching itself came from Paul Baran and Donald Davies. Cerf's genius is best described as architectural and social rather than theoretical: he did not discover a new principle so much as choose, correctly and early, the simplest set of rules that a global system could be built on, and then spend a lifetime defending them.

Legacy

Almost every digital act performed anywhere on Earth passes through protocols Cerf co-designed. He is one of the very few engineers whose work is universal and whose name is attached to it. He still wears the three-piece suits, still writes a column called "CERF'S UP," and still drives a car with the licence plate CERFSUP.

Achievements

• Turing Award — 2004
• Presidential Medal of Freedom — 2005
• Notable work: Transmission Control Protocol
• Notable work: Internet protocol suite
• Affiliated with Google, IBM and Stanford University
• Educated at Stanford University, University of California, Los Angeles and University of California, Los Angeles
• Worked as computer scientist, engineer and university teacher