Spotlight on: The supercomputers of Bristol and Bath
The world’s supercomputers are the backbone to some of the most exceptional innovations in engineering and data, underpinning everything from smart city projects to modelling high-tech simulations.
Just a few years ago, the most impressive supercomputing power was equivalent to the power most of us hold in the palm of our hands in the form of smartphones and tablets.
Now the world’s fastest supercomputer, the Chinese Sunway TaihuLight, is capable of reaching a Linpack benchmark performance of 93 petaflops (a unit of computing speed equal to one thousand million million floating-point operations per second) which, if you’re not a computing geek, is a lot – we promise.
“High-performance computing is involved in all parts of daily life whether you know it or not”
However, it’s not just China that’s leading the way with impressive computing power. Given that the South West is a key hub for high-tech in the UK, it has its fair share of impressive super computers powering exciting projects across the region. So we’ve gathered some of the most exciting of them all for you to read about below.
Cray CS400 at CFMS
Most notable in the world of supercomputing is Seymour Cray who designed the first commercially successful supercomputer in 1964, the CDC 6600, then capable of handling three million floating point operations per second.
Of course, things have come a long way since then with Cray now a US giant that designs, manufactures, sells and supports some of the world’s most advanced supercomputer systems. Not only that but they chose Bristol to set up its European, Middle East and Africa (EMEA) R&D lab in 2016.
“Providing a technology platform that enables rapid innovation and supports the acceleration of engineering design capability is central to our vision”
And it wasn’t long before the first Cray supercomputer in Bristol started making an impact. The Centre for Modeling and Simulation (CFMS) at the Bristol & Bath Science Park was the first to benefit from the Cray CS400 cluster supercomputer which has over 80,000 processor cores, along with 8 GPU nodes which are used for highly parallel applications as well as for rendering and visualisation.
Sam Paice, Chief Operating Officer at CFMS, previously told us: “With modelling and simulation at the core of product development, providing a technology platform that enables rapid innovation and supports the acceleration of engineering design capability is central to our vision.
“It was paramount to select a technology partner at the forefront of high-performance computing, and with whom we can develop a collaborative partnership beyond the technology, and support our strategic vision”.
One clever project that has already benefitted from this innovation in supercomputing is the UK’s solar car racing team (pictured left) which used the computer to design and model a prize-winning solar car for a competition in Australia later this year.
BlueCrystal4 at the University of Bristol
Just a few months ago the University of Bristol powered up its latest supercomputer, the BlueCrystal4 (BC4), which was designed, integrated and configured by OCF.
The computer can deliver over 600 trillion calculations per second to over 1000 researchers across the campus – powering the latest research in everything from paleobiology, biochemistry, physics and molecular modelling to life sciences and aerospace engineering – its power putting it in the top 500 high-performance computers in the world.
The University’s Director of Advanced Computing Simon Burbidge, at the time of installation, commented: “High-performance computing is involved in all parts of daily life whether you know it or not, from the design of aeroplanes, drugs and electronics to financial trading, and it is central to the research activities of the university.
“We provide the hardware, software and tools that are needed to run the systems and make them usable to the people doing the research, including expert help in programming and optimisation to make their programmes run faster and better.”
YellowDog virtual supercomputer
Not all supercomputers need to be huge whirring machines requiring their own office and a team of maintenance staff. Oh no. YellowDog is the Bristol-based startup that’s taking a whole new approach to supercomputing – virtually.
“We like to think we’re effectively building the world’s fastest supercomputer”
YellowDog is an app that pays you for your computer’s spare processing power and offers it to 3D animators, giving them their own virtual supercomputer to dramatically speed up the time taken to render animations.
YellowDog’s founder Gareth Williams (pictured left) previously explained to us: “We believe there is already enough computer power in the world. But there is an imbalance between supply and demand.
“Through our unique technology, we securely harness and monetise under-utilised computer power, enabling organisations to deliver incredibly quickly and cost effectively, through leveraging this unlimited power on demand. We like to think we’re effectively building the world’s fastest supercomputer.”
Having recently won the prestigious Oracle Award for Excellence in Innovation during Oracle Open World 2016 it’s easy to believe that this unique technology could change the way we source supercomputing power for good.
- You may like: YellowDog’s virtual supercomputer wins the Oracle award for ‘Excellence in Innovation’
Bristol is Open (BiO) is a collaborative project between Bristol City Council and the University of Bristol that’s taking the city into the future – bringing internet of things tech, 5G networks and 3D visualisations of data – and connecting people with places and things. Effectively becoming a programmable city run by supercomputing power.
You can find out more about the project and its huge potential range of applications in the video below:
But such a huge project involving such large amounts of data needs a sophisticated and high-tech solution to get running and stay well maintained.
Bringing the networking technology expertise is Nokia Networks – with the company contributing a combination of funding, technical staff and Nokia Bell Labs expertise for several major technology trials over the next three years using the BiO high-speed network around the city.
Zeetta Networks, which is commercialising the technology, will install, maintain and support its NetOS software-defined networking (SDN) operating system on the BiO network to provide a multi-tenant, multi-technology software platform which will enable ‘Experimentation-as-a-Service’.
At the heart of the platform is Zeetta’s Network Provisioning Portal which allows service providers to acquire, on-demand, a ‘slice’ of Bristol Is Open’s physical network for the development of their own application or service.
- Find out more: Nokia joins Bristol is Open for smart city video analytics
- You may like: Bristol Is Open wins smart cities award
Isambard – the ARM supercomputer
Ever expanding, the power of supercomputers is seemingly endless with a race to build the biggest ever present. Researchers at GW4 Alliance (Bristol, Bath, Cardiff and Exeter Universities) along with the Met Office and super computer maker Cray are now attempting to build the world’s largest ARM-based supercomputer.
If a success, the £3million project will see a build that boasts 10,000 64bit ARM cores – the largest production system in the world.
Project Leader Simon McIntosh-Smith, Professor of High-Performance Computing at the University of Bristol, said: “Isambard is an exciting experiment. If we discover that ARM processors are competitive in HPC, then Isambard could be the first of a new generation of ARM-based supercomputers, ushering in an era of wider architectural choice, with greater opportunities for differentiation between supercomputer vendors.
These outcomes should mean that scientists can choose systems more highly optimised to solve their problem, delivering even more exciting scientific breakthroughs at greater cost effectiveness than ever before.”
- You may like: Bristol startup tops $60m investment for AI chip