Enhance Reliability & Efficiency in Large Data Centres
A Reliability-Aware Resource Allocation Method in Disaggregated Data Centers
Tags: City University of Hong Kong, Hong Kong, Computing Technology
Data centers (DCs) are essential for internet services, yet their server-centric architecture limits resource efficiency and upgradability. The technology introduced here proposes a reliability-aware resource allocation method for DCs, optimizing resource use through disaggregation, enabling independent upgrades and scaling. It addresses the need for frequent hardware refreshes and improves resource utilization, offering a cost-effective solution for cloud providers like Microsoft Azure and Amazon Web Services. This approach promises enhanced service reliability and operational efficiency in large-scale DCs.
IP Type or Form Factor: Patent Pending; Process & Method; Software & Algorithm
TRL: 4 - minimum viable product built in lab
Industry or Tech Area: Computing Architecture; Computing Databases & Storage