NSF CSR Program Funds Gandhi's Cloud Computing Research

 

Anshul Gandhi, an Assistant Professor in the Department of Computer Science in the College of Engineering and Applied Science, has been awarded a Computer Systems Research (CSR) award by the National Science Foundation (NSF). The funding will allow him and his team of students to conduct research on cloud load balancers and their performance.

Professor Gandhi, who is also an affiliate professor in the Department of Applied Math and Statistics, an affiliate of the Smart Energy Technology Cluster, and leads the Performance Analysis of Computer Systems (PACE) Lab, will receive nearly $400,000 over the next three years for his research project titled, Scalable, Heterogeneity-Aware Load Balancing.

CSR funded-projects are notable because of the significant progress expected in a short period of time. NSF’s CSR program “supports transformative scientific and engineering research leading to the development of the next generation of highly performant, heterogeneous, power-efficient, environmentally sustainable, and secure computer systems,” according to the NSF.

In summarizing the research, Gandhi stated, “I would say the ultimate goal is to improve the design of load balancers and schedulers in today's inherently heterogeneous cloud computing platforms.” Load balancers distribute incoming requests in cloud based services, which are typically online rented services from companies such as Google, Amazon, and Microsoft.  However, there is variability between rented services, even if two identical services are purchased. This leads to an impact in performance because the load balancers will still treat these services as identical even though they are not. Differences arise due to different generations of a particular device or machine, or because of different resource utilization levels of the underlying physical servers. If the load balancers are able to identify the differences between cloud computing resources or services, they can provide better performance by adjusting the load distribution accordingly. This points to the need for load balancers for heterogeneous resources, which is the focus of Professor Gandhi’s research.

About the Researcher

Professor Gandhi earned his PhD in 2013 from the Computer Science Department at Carnegie Mellon University. He then spent a year as a post-doctoral researcher in the cloud computing division of the IBM T. J. Watson Research Center, before joining the faculty at Stony Brook University in 2014. Since joining Stony Brook, Gandhi has been awarded an IBM Faculty Award, a Google Faculty Research Award, and a Microsoft Azure Research Award.