Kaicheng Zhang's Experimental Analysis Highlighted by The Computation Institute

New research led by EECS PhD Student Kaicheng Zhang has been featured in an news article, titled, "A Machine Learning Cooldown for the Data Center" by The Computation Institute (CI). The joint project with Prof. Seda Ogrenci-Memik, Prof. Gokhan Memik, and Kazutomo Yoshii (Argonne National Labs) tested whether smarter task placement, the assignment of computing jobs to specific servers in a cluster or data center could successfully reduce system temperature and the need for cooling. Using Chameleon as a test environment, the group explored how machine learning can help make decisions about task placement that conserve energy without sacrificing performance.

Zhang is in his fifth year graduate studies, co-advised by Prof. Ogrenci-Memik and Prof. Memik, as well as a member of the Ogrenci-Memik Lab. His research interest is in improving the performance and energy efficiency of computer architectures, including micro-processors and DRAM memory.

The Computation Institute (CI) was established in 2000 as a joint initiative between The University of Chicago and Argonne National Laboratory to advance science through innovative computational approaches. Scholarship in the sciences, arts, and medicine depends increasingly on collection and analysis of large quantities of data and detailed numerical simulations of complex phenomena. Progress is gated by researchers’ ability to construct complex software systems, to harness large-scale computing, and to federate distributed resources. The CI is both an intellectual nexus and resource center for those building and applying such computational platforms for science. As an intellectual nexus, it brings together researchers from different disciplines with common interests in advancing the state-of-the-art in computing and its applications. As a resource center, it provides expert assistance to scholars whose work requires the most advanced computational methods.