Title :
Using platform level telemetry to reduce power consumption in a datacenter
Author :
Vincent, Jay L. ; Kuzma, John
Author_Institution :
Intel Corp., Folsom, CA, USA
Abstract :
The modern datacenter has significantly changed the way we need to consider managing the resources of power, space, cooling and compute. The traditional methods based on static deployments is no longer adequate. In order to address this transformation, a dynamic management methodology is required. This methodology can be greatly enhanced through the use of platform level telemetry. This paper will review the underlying technology instrumented in x86 based server platforms that enables real time monitoring of power, thermal and compute and how this data can be used to make optimization decisions.
Keywords :
computer centres; cooling; decision making; file servers; power aware computing; telemetry; computing resources; cooling resources; data center; dynamic management methodology; optimization decision making; platform level telemetry; power consumption reduction; power resources; space resources; x86 based server platforms; Computational modeling; Cooling; Optimization; Resource management; Sensors; Servers; Telemetry;
Conference_Titel :
Thermal Measurement, Modeling & Management Symposium (SEMI-THERM), 2015 31st
Conference_Location :
San Jose, CA
DOI :
10.1109/SEMI-THERM.2015.7100130