For Black History Month, CSE Spotlights Faculty and Alumni in Academia

  Bookmark and Share

Marcus Darden | Odest Chadwicke Jenkins | Jason Mars
Kyla McMullen | James Mickens | Kunle Olukotun

Kunle Olukotun

Kunle Olukotun (PhD CSE '91) is the Cadence Design Systems Professor of Electrical Engineering and Computer Science at Stanford University. His current research projects focus on domain-specific languages (DSLs) and specialized architectures.

Kunle Olukotun

Born in London, Olukotun was captivated by his math and science classes, ultimately leading him to pursue a career as a computer scientist – a stark difference from his father who was a lawyer and his mother who worked as a secretary. At the age of 12, his parents decided to permanently move back to their home country of Nigeria, where he continued to foster his knowledge of math, science, and technology. After completing secondary school, he convinced his parents that he should attend an American university to pursue his education in engineering.

Olukotun decided to attend Calvin College in Grand Rapids, MI, with the expectation of transferring to U-M after his junior year. At that time Calvin did not have a four-year engineering program. He joined the undergraduate electrical engineering program that was still on central campus in what is now East Hall. It was a great time to be at U-M with lots of innovative high-performance computing research being developed in the Advanced Computer Architecture Laboratory (ACAL). His desire to contribute to this exciting research led him to stay at U-M for graduate school. He completed his dissertation on high-speed Gallium Arsenide microprocessors under the guidance of his advisor Prof. Trevor Mudge, and earned his Ph.D. in Computer Science and Engineering.

Once he completed his Ph.D., Olukotun had a job lined up at Motorola in Austin TX, but an unplanned conversation with a Stanford professor, named Michael Flynn, influenced him to reconsider his future. It’s after this conversation in 1991 that he decided to apply for an assistant professor position in the EE department at Stanford University.

While at Stanford, Olukotun realized that microprocessors of that time wouldn’t efficiently scale up by making single processors more complex. To solve this problem, Olukotun and his students designed the first general-purpose multicore CPU, called Hydra. This processor was constructed by building four copies of a simpler CPU core on one chip instead a single complex CPU. The Hydra design, while much more scalable, forced programmers to face the difficult task of writing parallel programs with threads of instructions that must be executed together. The Hydra design was very controversial in the mid-nineties, when it was conceived, but now every cell-phone microprocessor uses this design approach.

Olukotun took his findings from Hydra and founded Afara Websystems, a company that designed and manufactured low power server systems with chip multiprocessor technology. He wanted to address the problem of data centers running out of power and space, and he believed it could be solved with multiple processors on a chip in conjunction with multi-threading. The Afara technology became known as Niagara. In 2002, Afara was acquired by Sun Microsystems, Inc., a company that manufactured computers, computer components, and software. Niagara became the center of Sun’s throughput -computing initiative and became the foundation for a line of popular computing systems. Sun was acquired by Oracle in 2010, and the Niagara line of processors is now in its ninth generation.

After returning to Stanford from Sun, Olukotun decided to tackle the problem of making it easier to develop parallel software for multicore processors. He became the director of the Stanford Pervasive Parallelism Lab (PPL), which seeks to make it easy to develop parallel programs in all application areas using Domain Specific Languages (DSLs). A big focus of his current research is on the problem of developing programming environments and specialized parallel hardware for big data analytics and machine learning. One of the goals of this research is to enable truly smart mobile devices. While at Stanford, Olukotun has advised over thirty PhD students, one of his well-known advisees is our own Prof. Valeria Bertacco. Olukotun is an ACM Fellow and an IEEE Fellow for his contributions to the field of computer architecture.

Posted: February 3, 2017