April 26, 2018 | Written by: IBM Research Editorial Staff
Categorized: Hartree Center
Share this post:
Chancellor of the Exchequer, George Osborne opening Hartree centre on 1 February 2013.
Five years ago, 1 February 2013, the Science and Technology Facilities Council’s (STFC) Hartree Centre was formally opened, bringing together a critical mass of computational scientists and engineers from STFC and IBM to reap economic benefits from advanced computer architectures and address important challenges for commercial businesses. Kirk E. Jordan, IBM Distinguished Engineer and Chief Science Officer, UK, IBM Research, who has been involved since its inception, reflects on five years of innovation in high performance computing and data science.
Early on, the focus of the Hartree Centre was to exploit high performance computing (HPC) for industrial benefit. The work horse systems were based on several IBM Blue Gene supercomputers and there were a dozen research projects including; next generation weather and climate algorithms, dissipative particle dynamics (DPD), computational fluid dynamics (CFD), consumable HPC for chemists; wheat genome data set and insurance risk analysis using disaster modelling such as earth quake rupture or hurricane simulations.
Just a year after the opening of the Hartree Centre, the HPC world was changing from an emphasis on floating point computation to bring more focus on data, which is why IBM put a focus on data centric solutions. In the UK, IBM Research and STFC joined together to take the data centric solution one step further to encompass cognitive computing.
Here we are five years later and the Hartree Centre and IBM are continuing to deliver on the original vision. The research collaboration has grown from a few research projects between teams separated by the Atlantic Ocean to bring more than 26 IBM Research staff based at Daresbury Laboratory connecting technologies from IBM’s global labs with STFC staff.
Together STFC Hartree Centre and IBM Research are currently working with 18 UK companies, e.g. Unilever, EDF Energy and many institutes, e.g. Rothamsted Research, to demonstrate that data centric cognitive computing can impact on their competitiveness. We expect to add an additional two dozen firms and organizations to this list by the end of the year.
Can Big Data and machine learning help farmers?
The work being done encompasses a variety of proof of concepts. For example, in chemistry, we have shown that using data centric cognitive computing we can reduce the number of simulations required by a third and improve the phase boundary curve resolution by four orders of magnitude. This is leading to improve time to market for companies making laundry detergent, shampoos, and oil additives, to name a few.
The benefit to the UK of our collaboration is the ability to bring in IBM Research’s background in intellectual property to bear on UK companies and institution’s challenges. One example of this is applying the decades of work on Physical Analytics and Integrated Repository and Services (PAIRS) and applying it to protect crops from moths and is in collaboration with with Rothamsted Research Institute. It’s being adopted for a systems-based approach for agricultural researchers, bringing together many different datasets and enabling them to carry out analytics and modelling easily.
And more recently, the team is working with a UK company and applying AI to improve home care for the elderly.
Five years later we are seeing that companies and institutes in the UK are realizing the Hartree Centre brings a critical mass of computational expertise exploiting both advanced computing architectures and developing new algorithmic approaches when coupled with the institutions’ expertise in their products, changes the way we all look at and find solutions. This is creating new research areas and demonstrating the investment return on research.
I personally believe that the public-private partnership of the Hartree Centre, working with a wide variety of industry collaborators, is bringing to fruition excellent research motivated by industry problems as conceived in the early day of the Centre five years ago, and reminds me of a saying that Colonel Hannibal Smith, from the 1980’s popular TV show the A-Team, would say, “I love it when a plan comes together.”