Innovation Watch: Computer memory designed to save the world

A team of scientists at Lancaster University have designed a new type of computer memory which could solve the digital technology energy crisis. We look at just how this memory upheaval works and the impact it could have.

Energy-saving bulbs, smart standby modes and turning the heating down. They’re all things we’re more conscious of in our bid to lessen the huge impact we’re having on the climate. Unfortunately, the power saved by these methods has been totally eclipsed by computers, rendering their benefits all but useless.

New research shows that by 2025 data will account for a fifth of global electricity. That’s huge! And it shows absolutely no sign of slowing down. The same research warns that it could account for 14% of global emissions in 2040.

The obvious conclusion here is that something seriously needs to be done. Something tells us that’s not going to be ‘use less computers.’

Luckily a new type of memory has been invented (and patented) by a team at Lancaster University. Typically, computer memory has to choose between being stable, slow and long term or prioritising speed over these traits. This new design uses quantum mechanics to solve this dilemma. Just don’t ask us to explain exactly how it does that, OK?

This ultra-low energy memory could replace the current standard Dynamic Random Access Memory (DRAM) industry, worth around $100bn. While DRAM is fast data constantly needs to be refreshed, which requires a power to do so. Removing that weakness could instantly reduce power consumption in data centres by a fifth.

This memory could also allow computers to instantly go into a power-saving standby mode, even between button presses, without the user knowing.

It feels like changing habits, or deciding what counts as a ‘worthy’ use of computers, is going to be a tough battle to win. If engineers and designers can create products built around solving issues like our energy crisis then they’re so much more likely to have an impact.

Also, from a selfish point of view: fast memory that’s reliable? Count us in!