'Computing's Nobel Prize' Winners Paved the Way to Smartphone Chips

Stanford's John Hennessy, now chair of Google parent Alphabet, and Berkeley's David Patterson developed the Reduced Instruction Set Computer in the 1980s.
Image may contain John L. Hennessy Tie Accessories Accessory Coat Clothing Overcoat Apparel Suit Human and Person
Alphabet Chairman John Hennessy founded MIPS and served as Stanford University president for 16 years.Linda A. Cicero/Stanford News

As a young computer-science professor at Stanford in the early 1980s, John Hennessy helped pioneer a new computing concept called RISC, for “reduced instruction set computer.” He expected others to see the value of the idea and commercialize it. When no one did, Hennessy started a company called MIPS.

"We never intended to be entrepreneurs," Hennessy says. "I thought what we were doing was so compelling our friends in industry would implement it. But there was too much of a 'not invented here' syndrome. That's why we ended up doing it."

Across San Francisco Bay, David Patterson was working on similar ideas at the University of California Berkeley. He later helped Sun Microsystems turn his team's work into its flagship SPARC line of chips.

Wednesday, Hennessy and Patterson were awarded the Turing Award, computing’s equivalent of a Nobel Prize for their work on RISC. The two didn’t invent RISC themselves. They and their teams built on existing research, but they popularized the ideas and helped prove they were feasible, and Patterson's team coined the term RISC.

Patterson describes computer architecture as the vocabulary that hardware and software use to talk with each other. Before RISC, he says, those architectures tended to be highly complex and full of the digital equivalent of polysyllabic words. What he and Hennessy did was essentially streamline the vocabulary to use only monosyllabic words.

David Patterson helped Sun Microsystems turn his team's work into its flagship SPARC line of chips.

Weinberg-Clark Photography/Association for Computing Machinery

This streamlined architecture quickly found a home in embedded devices, servers, and, until Apple switched to Intel processors in 2005, Macintosh computers. Today RISC is best known for powering most smartphones, thanks in large part to the chip company Arm Holdings, designers of the line of chips called ARM, short for Advanced RISC Machine. The company’s chips are also found in connected cars, drones, and a wide variety of gadgets that need low-energy processors.

Hennessy and Patterson didn't start out collaborating. But the pair say their papers influenced each other, and they later teamed up to author a textbook called Computer Architecture: A Quantitative Approach, published in 1989. The approach detailed in the book, which is still used in computer architecture classes today, was possibly even more influential than their RISC research.

Patterson describes the textbooks that came before as "catalogs" that simply explained what different types of computers did, but didn't provide any unified design principles. With their book, Patterson and Hennessy established a set of principles for evaluating the performance of different architectures, taking much of the guesswork out of designing computers. Forest Baskett, a partner at the venture capital firm New Enterprise Associates and former Stanford computer science professor, says Hennessy and Patterson established computer architecture as a engineering discipline and influenced chipmakers like Intel that don't make RISC processors.

Krishna Palem, a computer science professor at Rice University, describes their innovation as a new way to think about processors, one that made it easier to think about how the chips are designed. This new thought process made it cheaper to develop chips, because they required less extensive testing, and enabled designers to make more energy efficient chips. The lower costs meant that chip design was no longer the sole domain of big companies like IBM, and the increased efficiency meant that small, battery-powered devices could be imbued with computing power.

Patterson returned to academia after his stint at Sun, became chair of Berkeley’s computer science department, and later joined Google as a distinguished engineer in 2016. He’s also the vice chair of the board of the RISC-V Foundation, a coalition dedicated to building an open source processor architecture backed by companies including AMD, Google, and Qualcomm.

Hennessy ran Stanford's Computer System Laboratory from 1989 until 1993, then became chair of the computer science department, dean of the university's school of engineering, and Stanford’s president from 2000 to 2016. Along the way he also joined the boards of Google and Cisco. In February, he replaced Eric Schmidt as chair of Google’s parent company Alphabet.

Some critics argue that Stanford become too cozy with Silicon Valley during Hennessy’s tenure as the university’s president. But Hennessy points to his own experience with MIPS as an example of why academics should partner with private industry. If he’d simply published his work and waited for companies to build on it, RISC might have developed more slowly. “I think when we do work that has societal value, there’s a responsibility to get that out into the world, whether that’s cancer research or information technology,” he says.

In the Chips
  • Last year's winner of the Turing Award was Tim Berners-Lee, who created the technology that underpins the web.
  • Arm Holdings proved so successful that Softbank acquired it for $32 billion in 2016.
  • John Hennessy, David Patterson and Harvey Mudd College President Maria Klawe co-wrote this rebuttal to James Damore's memo about women and tech.