Singapore’s latest supercomputer, way more powerful than anything within the region, is transforming research and industry here. The machine, only a 12 months old, is already getting used to design safer, cheaper megastructures and to know how human genes cause disease. It has the potential to remodel research in lots of other areas, too.
The facility, called Aspire 1, is the center of the National Supercomputing Centre (NSCC) in Singapore, which was jointly established last 12 months by the Agency for Science, Technology and Research, Nanyang Technological University (NTU), the National University of Singapore (NUS) and the Singapore University of Technology and Design.
Located on the seventeenth floor of certainly one of the Fusionopolis towers in Buona Vista, the Fujitsu machine has greater than 30,000 cores, compared with two or 4 cores in a laptop computer. It can perform 1,000 trillion floating-point operations — decimal calculations — per second.
Last November, the Aspire 1 was ranked a hundred and fifteenth on the planet by TOP500, a widely respected supercomputer rating. By comparison, the world’s fastest supercomputer, China’s Sunway TaihuLight, has greater than 10 million cores and purrs at about 100 times faster than the Aspire 1.
But nothing in Southeast Asia compares to Singapore’s latest achievement.
It even inspired a brand new biennial complement to the magazine, Supercomputing Asia, which debuted here earlier this month, published by parent magazine Asian Scientist. One of the primary users of the Aspire 1 is Professor Allan Ross Magee’s group from the NUS Faculty of Engineering, who’re using it to assist design floating tanks for storing oil in a project supported by JTC Corporation.
The supercomputer allows them to more accurately simulate the complex movements of multiple oil tanks, affected each by the water currents around them and by the oil circulating inside them, across space and time, allowing them to refine their design, making it safer and cheaper.
Prof Magee said: “It allows us to see smaller-scale structures that affect physics in ways we didn’t expect… the design, how it’s built and how it’s designed are all dependent on computing power. “A smaller computer would give a less optimal solution within the time constraints of the project. Then we have to ‘redesign’ the system, which increases the cost.”
Another early adopter of the Aspire 1 is NTU Professor Stephan Schuster, who leads the international GenomeAsia 100K project, which goals to investigate the entire DNA sequences, or genomes, of 100,000 Asians. The nonprofit consortium hopes to make use of the knowledge to advance medical advances for Asian populations.
His team has an enormous task in front of the pc: they have to collect 100 billion statistical data points concerning the genome of everyone and compare them with the database.
Shortly after the Aspire 1 experiment was launched early last 12 months, scientists ran it at full power, using genetic data from 2,000 people at a time.
“We could give them (NSCC) a very large computational problem. We were a stress test for their infrastructure,” said Prof. Schuster. “When you stress the system, you can see how robust it is. Is the power management working, is the thermal management working, is the data transfer rate where it needs to be?”
Despite initial problems with data storage, the pc served its purpose.
In fact, the two,000 genomes used to check the pc have already yielded latest discoveries.
The data make clear the genetic diversity of hunter-gatherer populations in Asia, which Prof. Schuster believes is value preserving not just for cultural heritage but in addition to potentially overcome future challenges, similar to infectious disease outbreaks.
While powerful, the Aspire 1 is absolutely only sufficient for the needs of a single university the dimensions of NUS or NTU, noted Prof Schuster. As soon as the pc became available last 12 months, requests for its use went “skyrocketing” from each academia and industry, he added.
He estimates that the supercomputer’s capability ought to be increased 10-fold over the subsequent five years. For him, there is no such thing as a doubt that that is the suitable path. Using business computing services for his genome project would cost several million dollars.
“Without the NSCC we wouldn’t be able to do what we do now… there would be no project.”
Source :Strait Times
