The IRR is engaged across a broad range of Big Data activities. The Big Data risk analysis research programme is performed in conjunction with RSSB which represents the interests of 70 train operators, rolling stock owners, infrastructure providers and suppliers in the UK. Projects include vehicle and track monitoring, analysis of safety databases, risk identification and safety operation investigations.
Following a comprehensive market evaluation, the IRR decided to deploy Iceotope's PetaGen 1C high-performance computing system, in which Intel server electronics are immersed in a safe, non-flammable liquid coolant, to provide a dedicated computing resource for these activities.
By using liquid cooling, Iceotope's portable computer system operates in virtual silence and can deliver up to 7.5kW of computing power using blades housed in a cabinet occupying less than one square meter of floor space, without any requirement for bulky air conditioning units or acoustic hoods. The implementation supports the IRR's computing strategy, enabling high-performance computers to be located in the same facility as its railway research team for the first time.
Trials by early adopters of Iceotope's liquid cooling technology have demonstrated its ability to reduce cooling energy consumption by 80 percent compared with air cooling and achieve Power Usage Effectiveness (PUE) of less than 1.1. Further efficiencies have been achieved by organisations that reuse the waste heat captured by Iceotope's cooling system to heat facilities, further reducing energy bills.
Dr. Coen van Gulijk, Professor in Railway Safety and Risk at the Institute of Railway Research, commented: "Big Data analysis has huge potential to improve railway safety, but it requires significant computing resources. With its small footprint, low power consumption and silent liquid cooling, we can locate Iceotope's computing system alongside the research team in the laboratory, providing a dedicated, highly efficient high-performance resource."
Peter Hopton, founder and technology director at Iceotope, commented: "Big Data has created huge opportunities but requires greater processing power which generates a significant amount of heat. Traditional cooling approaches require huge amounts of energy and up to four times more real estate to accommodate air conditioning. We are delighted the IRR has adopted our pioneering multi-stage liquid cooling system. This deployment was down to the tremendous space, energy and cost savings that can be achieved using Iceotope technology and our ability to locate high-performance computing resource outside the data centre, near those that need it."