Joint Fujitsu-A*STAR Computational Bio-Medicine Forum

Ladies and Gentlemen, distinguished guests from Fujitsu Japan, Guests and Colleagues, It is my pleasure to welcome you to this first joint Fujitsu - A*STAR Computational Bio-Medicine Forum. Scientists will increasingly rely on computing to extract useful knowledge (not just information) from vast data resources. This will challenge the way scientists traditionally pursue scientific inquiry to gain insight and develop theories. What’s more, science will increasingly be driven by through work done on databases, finding relationships, developing models , testing simulations and visualization. And this means that scientists will increasing need to know computer science in much the same way our professional forefathers learnt mathematics. A*STAR, and IHPC in particular, have been building a new kind of relationship with Fujitsu Japan, through Fujitsu Asia representative offices in Singapore and directly with Fujitsu Japan, for the last two years. Our exploratory discussions led to the Memorandum of Understanding that was signed in January this year. Our MoU provides us with a concrete plan of action and articulates our commitment to this partnership. One of the important action items was establishing a strong joint R&D program with A*STAR and Fujitsu as the two partners, based here in Singapore. One clear goal of this program is to identify and conduct computational research of unprecedented scale in size, computational speed and complexity. We plan to study and solve very large computational problems arising from important branches of industry, engineering and science. What characterises the problems that will be subject of our joint research is not of immediate practicality and applicability but also scaleability. There are currently only three supercomputing facilities in the world with the computational capability of about 1 PetaFlops. Or an astonishing one billion operations pers second ! Fujitsu has an ambitious plan, and clearly the resolve, to build in two years (or less) a machine that will be 10 times more powerful than the largest one used today.Fujitsu’s FX-1 machines, based on Fujitsu’s own sparc VIII-fx chip, the fastest chip in the world, have already been installed in at least three research establishments in Japan. The new Fujitsu 10 PetaFlop supercomputer is planned to evolve from the current FX-1 design. ASTAR-Fujitsu joint computational research is aimed at reaching the computational power limits of the supercomputer that is being now designed in Fujitsu. A*STAR has the breadth of expertise and repertoire of scientific pursuits to challenge the prowess of the biggest computing systems. In fields like genomics, large dimensional data visualization, imaging, large scale complex engineering systems there are plenty of opportunities to develop new theories and deeper understanding of scientific phenomenon and engineering systems. Today’s symposium is a demonstration of potential areas that match scientific interests to computing resources. The idea of the symposium, where a select few areas in computational biology and medicine will be presented, arose from a workshop between IHPC and Fujitsu scientists in Tokyo, in May this year. During our discussions and presentations it became apparent that both Fujitsu and A*STAR are involved in research topics in computational biology and medicine that closely overlap each other with a high degree of synergy. Naturally, within the limited time of a one day forum, we can only feature a few selected topics in the burgeoning domain of computational biology and medicine that is representative of our interests and achievements. I am sure there will be other areas of common interest as we advance in our partnership. The topic of this morning’s session is “The heart in a machine” – or in other words, how we model the human heart and how through this representation of the geometry, shape, muscle structure, hydrodynamics of blood flow and bio-molecular processes driving the heart action, our insight and also the potential for early detection and treatment enhances. The afternoon session offers some disparate topics. We will hear two talks about in-silico drug design and rapid screening of drug leads. We will learn about advances in DNA Aptamer development, and possible service. We will get a glimpse of how Nature provides us with ideas for very practical engineering solutions through new methods of delivering drugs in the form of vaporisers. Finally we will learn about understanding multiscale biological processes and structures from the molecular to the cell level. Here a team from BII run their codes for 80 days non-stop in order to understand complex, very large molecules and assemblies at sub-millisecond time periods. I am sure you will learn new things from our excellent group of speakers today. Much like the time where sophisticated astronomical tools helped astronomy advance through the golden age of space research; I feel we are at the cusp of a similar golden era of computing being the “magic carpet” that helps advance engineering and biological science. I wish you all a very pleasant and productive meeting. Thank you. Dr Raj Thampuran