Today I spoke at the Data Centre GreenTech conference in Sydney. For once an IT infrastructure conference not withering on the vine. Plus, a broad range of content that smells innovative, and an overall attitude of ‘lets do better’ versus ‘lets sell you some IT shit’!
Some amazing stats from Kenneth Brill, arguably the godfather of PUE. Brill states, in the US, data center energy consumption (as a percentage of the overall grid) has increased from 1% in 2000, to 3% today, and is projected to 10% in 2020. Brill is an amazing guy – basically defined many operating standards for data centre efficiency and tier rating/design as we know them today. He clearly recognises the need to progress beyond the PUE centric thinking he incubated 10-15 years ago…
My talk suggested a framework for ‘Sustainable Data Centre Energy Management’, with a focus on closing critical gaps in the industry’s ability to manage data centre energy effectively and in a more sustainable way (i.e. not uncontrolled growth). The basic concept is to apply storage (asset) management disciplines to energy management, with a focus on measurement, process, governance, and putting real measurement in place.
One low-hanging fruit is to advance well and beyond PUE as a measurement stick for data centre energy efficiency. Good timing, since PUE has now attained global harmony for data centre energy efficiency measurement!
We’re suggesting CPU efficiency (KW/h divided by # of CPU cores) and Green House Gas (GHG) quantification as more meaningful metrics, in addition to PUE/DCIE, which are relative but easily skewed metrics. It doesn’t have to be rocket science as the next step beyond PUE. There are a dozen other practical and measurable metrics that for some reason are lost on the industry as the ‘next generation measurements’ transport me back in time to falling asleep in 3rd year calculus to an instructor who couldn’t speak English!
It’s good to be involved in a surprisingly fresh movement around data centre efficiency. The overall attitude is constructive and people are trying out new ideas. Reminds me of why I didn’t take the job on a rig in the Gulf of Mexico back in 1996.