According to The Independent, the power used by data centres accounts for approximately 3% of the global electricity supply – and with the rate at which data is being created, this percentage is only set to rise.
Now more than ever, data centres should be doing whatever they can to minimise the impact of their energy use. However, hyper-scale amounts of energy and the increasing popularity of web-scale IT systems mean that advancements in hardware generally seem to be at odds with the goal to become more energy efficient.
However, there have been improvements, such as making the best use of outdoor air and data centre design, as well as companies choosing site locations that make use of renewable energy. But the storage industry should always be looking for alternative ways to save energy – which is where artificial intelligence (AI) can play a part.
In October, Google announced that its DeepMind AI software reduced the cooling bill for the company’s data centres by 40%. This technology uses data such as temperature readings, power usage, and pump speeds to improve energy efficiency. Whilst this software may not be as effective for smaller data centres, it’s a step in the right direction.
More recently, researchers at Lancaster University created a self-assembly software system driven by AI. The machine-learning tool uses search and sort algorithms and memory caches to help it alter its own behaviour and carry out server tasks in the most efficient way.
This software can cut data centre running costs significantly, and could signal the dawn of a more efficient era for the data centre industry. With so much more data being created, artificial intelligence could help adapt to the changing demands of data centres, now and in the future.
At LDeX, we’re committed to exploring new ways to save energy – click here to find out more.
