CSE alum and computer architecture innovator Kunle Olukotun chosen for top recognition by ACM/IEEE
Michigan alumnus Kunle Olukotun (CSE MSE ‘87 CSE PhD ’91) has been selected as the recipient of the 2023 ACM-IEEE CS Eckert-Mauchly Award “for contributions and leadership in the development of parallel systems, especially multicore and multithreaded processors.” The Eckert-Mauchly Award is widely recognized as the computer architecture community’s most prestigious award.
Olukotun is the Cadence Design Systems Professor and a Professor of Electrical Engineering and Computer Science at Stanford University and is co-founder and Chief Technologist at SambaNova, a provider of generative AI platforms. SambaNova is his third startup.
In the early 1990s, Olukotun became a leading designer of a new kind of microprocessor known as a “chip multiprocessor”—today called a “multicore processor.” His work demonstrated the performance advantages of multicore processors over the existing microprocessor designs at the time. He included these ideas in a landmark paper presented at the ACM Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS 1996), entitled “The Case for a Single-Chip Multiprocessor.” This paper received the ASPLOS Most Influential Paper Award 15 years later. Olukotun’s multicore design eventually became the industry standard.
Olukotun’s insights on multicore processors and thread-level speculation research laid the foundation for his work on fine-grained multithreading, a technique which improves the overall efficiency of computer processors (CPUs). These designs were the basis for Afara WebSystems, a server company Olukotun founded that was eventually acquired by Sun (and later Oracle). Sun Microsystems used Olukotun’s designs as a foundation for its Niagara chips, which were recognized for their outstanding performance and energy efficiency. The Niagara family of chips are now used in all of Oracle’s SPARC-based servers.
Later, with Christos Kozyrakis and others, Olukotun was a leader in designing the Transactional Coherence and Consistency (TCC) approach to simplify parallel programming. He was a co-author of the paper, “Transactional Memory Coherence and Consistency,” which was presented at the 2004 International Symposium on Computer Architecture (ISCA) and received the Most Influential Paper Award in 2019. Olukotun is one of only two researchers who have received the Most Influential Paper Award from both ASPLOS and ISCA.
Olukotun’s work establishing the coarse-grained reconfigurable dataflow has played a pivotal role in machine learning and other data-intensive applications. In 2017, after publishing several papers on this approach, he co-founded the startup SambaNova.
Olukotun was advised by Trevor Mudge, the Bredt Family Professor of Engineering, during his graduate studies at Michigan. Today, Olukotun has authored more than 200 publications which have received over 20,000 citations, and he has been issued 12 patents. He is a member of the National Academy of Engineering, an ACM Fellow, and an IEEE Fellow. He also received the IEEE Harry H. Goode Memorial Award.