Nearly 26,000 kWh Was Used On Linux Benchmarking In 2016

For those curious how much power is consumed with our constant flow of benchmarks and working 365 days per year, 2016 rang in at 25,943 kWh.

The power bill arrived this weekend with the last numbers for 2016. From January through December the grand total ended up being just under 26,000 kWh, a far majority of which was used for our basement benchmarking server room and in particular cooling all of those computers plus all of the other systems used for the Linux/BSD/macOS/Solaris benchmarking done at Phoronix.

Year-over-year the power use is only a bit higher. In 2016 I did try changing the flooring for lower temperatures but it didn't end up having too much of an impact, but at least now the flooring is stronger and looks better while being easier to clean.

This winter I also continue to be using this ducting setup for piping warm server air into the rest of the home, albeit it's been a cold winter and that can't do the complete job of warming the house on those frigid days.

A few weeks ago I also phased out some of the older systems (like quad-core 2.3GHz Opteron systems that were drawing 125 Watts per CPU socket yet so slow by today's standards...) to lighten the power-load a bit and replace them by some extra and more efficient systems. I also lightened the workloads for some of the daily benchmarking systems of the tests done at LinuxBenchmarking.com and halted some of the trackers/systems that were no longer producing useful data.

If anyone comes up with any novel power conservative ideas for 2017, feel free to share your ideas in the forums. Hopefully still within a few years I'll be able to invest in a solar roof to offset all of our power-hungry benchmarks.