By David Turek
A major challenge in cardiology is to predict who will die suddenly from ventricular arrhythmias – the most common cause of sudden cardiac death, which itself is the largest cause of natural death in the U.S.
Despite years of intense medical research, likely victims are hard to predict and even if identified, there are not effective and low-cost therapies available.
Mathematical models have the potential to provide insight into the mechanics of arrhythmias and sudden cardiac death, but we’ve never had the computational power necessary to make a model run even close to the speed of a real beating heart. Instead, researchers have been forced to work at low resolution, settle for short run times of – at most – a few beats, or take hours for a single heart beat.
Enter Sequoia, an IBM Blue Gene/Q supercomputer built on the IBM POWER Architecture at Lawrence Livermore National Lab. The 16+ petaflop system, #2 on the TOP500 list of the world’s fastest supercomputers announced today is currently being used to run a code called Cardioid, which was created by IBM and LLNL researchers to realistically model a beating human heart at high resolution.
For the first time, fully detailed whole heart models can be run quickly enough to examine how potentially fatal arrhythmias develop and are influenced by individual genetic variations, the administration of drugs and the use of medical devices. The IBM, LLNL team envisions this capability could eventually be widely adopted by medical centers, pharmaceutical companies and medical device firms, helping them to study and better understand the mechanisms that can lead to heart ailments.
For example, an initial modeling study investigated a class of drugs designed to prevent arrhythmia, but later produced greater mortality in some patient populations. Drugs targeting many diseases have the unfortunate side effect of increasing arrhythmias and sudden cardiac death. Understanding the risk of arrhythmias remains a challenge for the both pharmaceutical industry and the regulatory agencies because the danger may only be revealed after a drug is administered across a large and diverse patient population.
The Cardioid simulation work has been named as a finalist in the 2012 Gordon Bell Prize competition, which annually recognizes the most important advances in HPC applications and will announce this year’s winner on Thursday.
Other notable IBM systems on the TOP500 listinclude the fastest supercomputer inEurope, the POWER Architecture-based Blue Gene/Q system “JUQUEEN” at the German research center Forschungszentrum Juelich, which will be made available to scientists via a peer review process for projects ranging from biophysics to plasma physics.
The fastest system inCanadais also a Blue Gene/Q. It will be applied toward, among other things, such projects as brain research, assessing how climate change impacts watershed and better predicting consumers’ energy needs. It will be used to help make significant contributions to scientific breakthroughs and also help small and medium-sized enterprises in Ontario speed product research and development.
The National Center for Atmospheric Research’s (NCAR) Yellowstone supercomputer, debuting on the TOP500 at #13, last month began supporting initial scientific projects on a wide range of Earth science topics that will improve predictions of tornadoes, hurricanes, earthquakes, droughts, fires and other natural hazards.
For all of these organizations, it’s important that research is completed with energy efficient, highly-utilized systems to lower costs, especially those that use public funding. Beyond the immediate economic challenges of unused assets and higher energy bills, also at stake is the long term ability of clients to innovate and grow. Of more than 300 global businesses surveyed by IBM and IDC, only 21 percent were operating highly efficient data centers, but those companies are able to spend 50 percent more on business innovation and growth.
Sequoia is 91 percent water cooled and 9 percent air cooled. This allows the system to achieve more performance, while simultaneously consuming less electricity. IBM Blue Gene/Q systems account for 25 of the 30 most energy efficient systems on the TOP500 list. Today, one of the fastest computers in the world is also one of the most energy efficient.