KU Engineering Professor Wins NSF CAREER Award for Datacenter Networking Research
An assistant professor at the University of Kansas School of Engineering is the winner of a five-year, $533,000 grant from the National Science Foundation for his research rethinking the internal networking of datacenters — work to reduce latency, cut power consumption and accelerate speeds, all by maximizing the capacities of computing hardware.
Mohammad Alian, an assistant professor in the Department of Electrical Engineering and Computer Science, is the recipient of an NSF Early Career Development (CAREER) award for his work on what is known as a Near-Memory Datacenter Network.
The plan is to create inter-chip optical interconnections that deliver data directly from top-of-rack switch ports to processor chips and memory units by following a single instruction. No longer needed would be the comparatively stifling combination of multilayered software stacks, complex network protocol processes, frequent data movements and ongoing device management common in current datacenters.
The potential network architecture would be transformative, able to deliver remote memory access at faster than 500 nanoseconds more than 99.9% of the time, all while using 100 times less communication energy. And in a data-driven world where bandwidth requirements for data centers double every 12 or 15 months, finding efficiency is critical.
“Thanks to the advancement in CMOS technology scaling, also known as Moore’s Law, we can build chips that are amazingly fast in processing the data — but if we cannot feed enough data to these chips, they are useless,” Alian said. “This award will enable us to re-architect the datacenter network, from the application down to the hardware layer, to deliver data between the extremely fast processing elements at speeds close to hardware limits.”
CAREER awards are considered among the NSF’s most prestigious, given to about 500 early-career faculty each year with the potential to serve as academic role models in both research and education. NSF expects recipients’ activities to build a firm foundation for a lifetime of leadership in integrating education and research.
Alian plans to use award financing to provide training for K-12 teachers as well as for undergraduate and graduate students, with a focus on underrepresented minority and first-generation students in rural Kansas. He expects such training to boost university enrollment in computing and to help address the shortage of information technology professionals across the state and nationwide.
Alian also looks for the award to help expand opportunities for technology transfer, including building upon existing efforts. His team finished second place in last year’s Samsung Memory Open Innovation Contest, which focused on technology to accelerate memory-intensive applications and won the team $40,000 for research and development.
The CAREER award adds even more momentum, confirming the importance and potential impact of Alian’s work.
“Every time that you use your browser, post on your social media or check your news feed, you are sending requests to a datacenter,” he said. “Datacenters are the backbone of our digital world, serving the needs of billions of users. To be able to operate future datacenters we need innovations in the datacenter network architecture, and this is what our research is about.”