Comments on: Core4 Launches Cooling Product Line http://www.datacenterknowledge.com/archives/2009/06/22/core4-launches-cooling-product-line/ News and analysis about data centers, cloud computing, managed hosting and disaster recovery Mon, 13 Feb 2012 17:24:17 +0000 hourly 1 http://wordpress.org/?v= By: Alex http://www.datacenterknowledge.com/archives/2009/06/22/core4-launches-cooling-product-line/comment-page-1/#comment-4329 Alex Tue, 23 Jun 2009 01:30:37 +0000 http://www.datacenterknowledge.com/?p=12103#comment-4329 I refer readers to the work of The Green Grid – PUE – the initiative to have a standard calculation for the ratio between energy used by the compute devices and other energy defined – the largest being cooling. A computing device is, from a heat perspective, a variable heater. As a general guide the server uses 50% of it’s energy doing no compute work (fans, memory etc) and the other 50% in direct proportion to the cpu utilization. You do not need to measure the temperature in a data center to know how much additional cooling is required. Simply take the starting temperature and measure the watts watts being delivered from the power supply system = the heat being created = heat to be removed. I write this to encourage data center managers looking at ways to save energy, to both look at the cooling systems, and to look at some of the open source software initiatives to measure and graph energy use. open4energy ( http://open4energy.com ) is an open source energy monitoring project built on Cacti and RRDtool, both proven platforms for graphing time series data. I refer readers to the work of The Green Grid – PUE – the initiative to have a standard calculation for the ratio between energy used by the compute devices and other energy defined – the largest being cooling.

A computing device is, from a heat perspective, a variable heater. As a general guide the server uses 50% of it’s energy doing no compute work (fans, memory etc) and the other 50% in direct proportion to the cpu utilization.

You do not need to measure the temperature in a data center to know how much additional cooling is required. Simply take the starting temperature and measure the watts watts being delivered from the power supply system = the heat being created = heat to be removed.

I write this to encourage data center managers looking at ways to save energy, to both look at the cooling systems, and to look at some of the open source software initiatives to measure and graph energy use.

open4energy ( http://open4energy.com ) is an open source energy monitoring project built on Cacti and RRDtool, both proven platforms for graphing time series data.

]]>