16 December 2016
To aid the accuracy and timeliness of weather forecasts, the Met Office has one of the world’s most powerful supercomputers at its HQ in Devon. Its 500 scientists carry out research using data-intensive, high-resolution models to increase forecast accuracy and provide a deeper understanding of climate change.
Founded in 1854, the Met Office is now part of the Department for Business, Innovation and Skills and has a total staff of around 2,000. Its purpose-built headquarters building, costing £80m in 2003, is near Exeter Airport, and it also has a forecasting centre in Aberdeen and offices in Gibraltar and the Falklands.
The Met Office selected three vendors to provide high-performance computing (HPC) capabilities for its new Scientific Processing and Intensive Compute Environment (SPICE) which is designed to cut the time needed to analyse massive amounts of climate simulation
It chose SGI to power its SPICE initiative and to upgrade its Managed Archived Storage System (MASS), a repository for the research data carried out on the supercomputer as well as global observational data. By 2020, MASS is predicted to grow to about 300 petabytes.
SGI, which installed DDN (DataDirect Networks) storage, says processing capacity has increased, furthering researchers’ understanding of meteorology on a global scale.
The Met Office also chose to integrate Bright’s Cluster Manager so that the hardware, operating system, HPC software, and users of SPICE can be administered from a single point. It also installed Bright OpenStack which is designed to enable easy deployment, provision and management of its OpenStack-based private-cloud infrastructure.
SGI says the Met Office can now scale SPICE storage predictably while delivering high throughput to handle simultaneous data reads/writes.
Researchers, it says, can spin up virtual machines easily and operate their own private virtual environments with full control and direct access to their local network. In addition, they can easily increase the capacity of the virtual environment by adding more servers.
SGI says scientists using SPICE have already noted significant performance advantages over previous systems, resulting in much quicker analysis to support ongoing research. It claims massive volumes of data are now analysed in hours rather than days.
