Any views expressed within media held on this service are those of the contributors, should not be taken as approved or endorsed by the University, and do not necessarily reflect the views of the University in respect of any particular issue.

【BAYES COFFEE HOUSE TECH TALK SERIES】Harmonizing Diverse Compute Resources in the Cloud for Efficiency

Boris Grot, a Professor in the School of Informatics at the University of Edinburgh, will give a talk on the efficiency of cloud computing. He will explore how virtual machines (VMs) can achieve both cost efficiency and elasticity with serverless instances, addressing the challenges of load fluctuations and sudden spikes in online services.

Title: Harmonizing Diverse Compute Resources in the Cloud for Efficiency

Speaker: Boris Grot | University of Edinburgh

When: 02/06(Thur) 11:00-12:00 (UTC+00:00)London

Where: Edinburgh Research Center MR 1

Internal: https://imeeting.huawei.com/meeting/joinwelink?id=99125086&pwd=Nzk2NjU1&token=YyIBxcSiI281Xk0zh0uuV5iSbaYWk7KWS&stype=0

External: https://app.huawei.com/wmeeting/join/99125086/YyIBxcSiI281Xk0zh0uuV5iSbaYWk7KWS

Meeting ID: 99125086

Passcode: 796655

Registation: https://www.smartsurvey.co.uk/s/3N8U7J/

Abstract:

Today’s online services deployed in the cloud are provisioned on virtual machines (VMs) under the belief that VMs provide both performance and cost-efficiency. But is the belief justified? I will demonstrate that online services tend to exhibit significant load fluctuation at fine-grained intervals, even when coarse-grained load measurements indicate a steady load. Under such circumstances, VM capacity must be provisioned for the peak steady-state load in order to meet the SLO, thus resulting in cost overheads given the underlying fluctuations. Sudden load spikes further exacerbate the VM provisioning problem, necessitating either expensive stand-by capacity to absorb the spike or sacrificing SLO and end-user experience while additional resources are deployed on-demand. In contrast to VMs, serverless computing is highly elastic but is prohibitively expensive for serving a large volume of requests. I will make the case for diversifying compute resources by judiciously combining VMs with serverless instances to achieve both cost-efficiency and elasticity.

Bio: 

Boris Grot is a Professor in the School of Informatics at the University of Edinburgh, where he leads the EASE Lab. His research interests include server hardware and software stacks, networking, and datacenter-scale computing. Boris is a member of the MICRO and HPCA Hall of Fame and a recipient of multiple awards for his research. Boris was the Program Chair for MICRO 2022, General Chair for HPCA 2024, and is an Area Chair for ISCA 2025.

css.php

Report this page

To report inappropriate content on this page, please use the form below. Upon receiving your report, we will be in touch as per the Take Down Policy of the service.

Please note that personal data collected through this form is used and stored for the purposes of processing this report and communication with you.

If you are unable to report a concern about content via this form please contact the Service Owner.

Please enter an email address you wish to be contacted on. Please describe the unacceptable content in sufficient detail to allow us to locate it, and why you consider it to be unacceptable.
By submitting this report, you accept that it is accurate and that fraudulent or nuisance complaints may result in action by the University.

  Cancel