Self-Assembling AI Data Centre Software Boosts Server Efficiency And Saves Energy

A Facebook data centre. Image credit: Facebook

Researchers at University of Lancaster’s Data Science Institute have created REx, a smart self-assembling software system aimed for use in data centres

Artificial intelligence (AI) may be the key to reducing the massive amounts of energy consumed by data centres through using smart software that dynamically adjusts to handle different tasks more efficiently.

Researchers at University of Lancaster’s Data Science Institute have created REx, a software system aimed for use in data centres, which can rapidly self-assemble and adapt to the variety of tasks data centre servers handle without human interaction.

Smarter server software

Art Robot © Kjpargeter Shutterstock 2012Using a technique called micro-variation, REx uses a large library of software components, such as memory caches and various search and sort algorithms, that act as building blocks for it to reassemble itself intelligently to enable the most efficient running of server tasks.

“Everything is learned by the live system, assembling the required components and continually assessing their effectiveness in the situations to which the system is subjected,” said Dr Barry Porter, lecturer at Lancaster University’s School of Computing and Communications.

“Each component is sufficiently small that it is easy to create natural behavioural variation. By autonomously assembling systems from these micro-variations we then see REx create software designs that are automatically formed to deal with their task.”

REx uses three complimentary layers to achieve its self-assembly traits. At the base level is a component-based programming language called Dana that enables REx to search, select and assemble the software building blocks. Above that sits an assembly and learning framework that perceives the behaviour of components and configures them to suite the task at hand.

The third layer provides an online learning process which helps REx figure out the best software configurations in real-time through a statistical learning method called linear bandit models; at its most simple a means by which to acquire knowledge and create a strategy through statistical data exploration.

The idea is that REx’s ability to adjust itself on the fly means it can be put in charge of servers handling different tasks and take over some of the processing they need to do, therefore reducing the amount of energy they consume.

AI-Powered Future

“As we use connected devices on a more frequent basis, and as we move into the era of the Internet of Things, the volume of data that needs to be processed and distributed is rapidly growing. This is causing a significant demand for energy through millions of servers at data centres. An automated system like REx, able to find the best performance in any conditions, could offer a way to significantly reduce this energy demand,” explained Dr Porter.

The next steps are to explore how REx can be used to make its own software components in order to push the automation process even further.

The use of AI-powered systems in data centres is not only to ensure servers can run more efficiently but to also cut down on the need for human programming and the need for large teams of developers to maintain complex software.

A self-managing system would not only help reduce the energy consumption of data centres but also help reduce the cost of maintaining large teams of software engineers; a boon for both the environment and data centre operators.

AI is likely the next big step for the IT world, as more autonomous software creation systems are evolving alongside REx. Google has managed to create an AI that can create its own encryption without human interference.

The technology industry is characterised by rapid change and populated by colourful figures. New developments are often so transformational they seem hard to believe… and in some cases natural scepticism is justified. But can you spot the fake stories from the real ones?