Smart grid redirects computing power to heat homes EU-funded researchers have developed a decentralised High Performance Computing grid that actually uses calculations to heat houses and businesses. Digital Economy © Qarnot Computing The EU-funded project can benefit end users by helping them to reduce energy bills and by spreading intelligent heating to homes, offices, schools and health care buildings. ‘The idea is that we sell High Performance Computing (HPC) processing power through the cloud,’ explains EeHPC (Energy efficient High Performance Computing) project coordinator Paul Benoit, CEO of Qarnot computing. ‘Calculations are then smartly distributed to a remote computing grid, which is composed of calculation units that are actually electric heaters. Heat is therefore produced by decentralised microprocessors performing the computing.’ The electricity consumed by these heaters, called Q.rads generates heat that can then be directly used by the end user. Benoit explains that this concept allowed him and his team to progressively build a scalable computing grid, avoiding the massive investments of a data centre. Through the EeHPC project, a newer, more efficient version of the digital heater has been designed and a pricing review has been undertaken. Seasonality has also been taken into account. Intelligent calculation units can be installed in buildings such as schools, which tend to be closed during the summer, in order to significantly reduce demand for heating. The computing grid will be able to guarantee a certain capacity all year long, with extra power offered for free to universities. Cloud computing is about sharing computing resources rather than having local servers or personal devices handling each individual application. In this sense, ‘cloud’ is a metaphor for ‘the Internet,’ where services — such as servers, storage and applications — are delivered to an organisation's computers and devices through the Internet. This means that high performance computing’s power to perform tens of trillions of computations per second can be tapped into for consumer-oriented applications. By providing cloud computing through a distributed infrastructure where computing power is no longer concentrated at data centres, Qarnot believes the energy-efficient and sustainable HPC platform it has developed will benefit both cloud computing clients, who will have access to a more scalable and cost-effective infrastructure, and the end user. ‘The growth of huge data centres for storage and calculation has enabled scientists and industry to stay ahead of the curve,’ says Benoit. ‘However, these clusters are highly energy-intensive, not only because of the number of calculations they make but also because of their immense cooling systems. With our innovation, there is no need for infrastructure cooling as the heat produced is directly used for heating purposes.’ The EU-funded EeHPC project gave Qarnot, the company behind the Q.rad innovation, the opportunity to refine its business plan in order to fully exploit the billion-euro HPC market, which is expected to grow annually by over 5 % from 2015 to 2020. The emergence of big data has increased demand for systems that can handle data-intensive workloads. In Phase II of the project, Qarnot plans to industrialise the new digital heater for high-end quality production and to implement a corporate-grade cloud HPC platform through remote computing services. For further information, visit the project coordinator website and on Twitter. Keywords EeHPC, high performance computing, HPC, Qarnot, cloud computing, heating, electricity, big data, grid, infrastructure
