Case Study: The University of British Columbia achieves up to 89.8% cost savings with Google Cloud Platform

UBC research team develops framework to save costs for serverless computing

The University of British Columbia, a top global research university in Canada, needed a more efficient way to manage serverless computing workloads and reduce cloud costs. Working with Google Cloud Platform, the team used Google Cloud Functions and Pub/Sub to build UnFaaSener, an open-source framework for dynamically offloading functions across available hosts.

Google Cloud Platform helped UBC create a solution that monitors resource availability, predicts capacity, and transparently shifts workloads to idle virtual machines or other hosts. The result was improved ease of use for researchers and developers, plus up to 89.8% cost savings in some scenarios, along with added support from the Google Cloud Research Innovator program.

University of British Columbia

Mohammad Shahrad

Assistant Professor of Electrical and Computer Engineering