Innovation
Verne Global ’ s Icelandic data centre can process , analyse and store data and optimise computing capacity while reducing carbon footprints
Europe . This includes the assembly of some of the largest and most complex genomes , including the 17GB wheat genome , which can take between six and eleven terabytes of memory per run . The Institute also hosts a large , highthroughput compute cluster with more than 4,000 processing cores . Researchers use this for downstream analysis , along with over seven petabytes of storage , including 64 terabytes of the latest Intel NVME Flash technology for demanding inputs / outputs ( IO ) critical tasks .
Earlham Institute selected Verne Global based on its previous expertise in providing long-term , low-cost , sustainable power for computing as well as Verne Global ’ s experience working with private and public organisations . Via NREN providers Janet and NORDUnet ,
Earlham Institute is the first research organisation to examine the benefits of migrating a strategic , collaborative bioinformatics analysis platform to the data centre campus in Iceland .
Options to scale Like any research institute that is governed by large data-driven science , Earlham Institute is constantly dealing with large volumes of data arriving at very high velocity . This puts significant strain on its computing storage infrastructure , requiring increased storage space and data centre hosting capability , as well as increased operational cost to cool the infrastructure . As the trend for HPC in scientific research continues to rise , the Education , Research and Life Science industries need options that can scale to meet demand and help lower the costs of power . If research institutes , and other enterprise companies leveraging HPC to better understand the mysteries of the world around us , want to overcome these challenges and spend more of their budget on the actual research and less on the power bill , then they need to adopt new approaches . And , as is apparent from the Earlham Institute example , moving latency-tolerant , power-hungry applications such as HPC to locations where they can utilise renewable power supplies makes both financial and environmental sense . Imagine how many needles can be found when power costs are reduced by 70 % and more resources can be applied to advancing research further and faster ? ■
1961
Derek
The history of big data
Price concludes in Science Since Babylon that the total number of new journals grows exponentially rather than linearly
1996
Digital storage becomes more cost-effective than paper for data , according to Morris and Truskowski in The Evolution of Storage Systems
1998
John
R . Masey , Chief Scientist at SGI , presents a paper titled “ Big Data … and the Next Wave of Infrastress ” at a USENIX meeting
1999
A panel titled “ Automation or interaction : what ’ s best for big data ?” is held at the IEEE 1999 conference on Visualisation
44 | UKSPA breakthrough | winter 2017