July 31, 2013
Officials at Australian National University (ANU) in Canberra today took the wraps off Raijin, a new 1.2 petaflop supercomputer that is fastest ever in the country. The new cluster is housed at the new National Computational Infrastructure (NCI) facility on the ANU campus, and will be used for climate modeling, among other uses.
Raijin is a Fujitsu Primergy cluster that features 57,472 Intel Sandy Bridge cores running at 2.6 Ghz, 160 TB of memory, 10 PB of storage, and a Mellanox FDR Infiniband interconnect that delivers 9 terabytes per second bandwidth, according to ANU.
The cluster debuted at number 24 on the Top 500 list in November 2012, and dropped to number 27 for the June 2013 list. The tests used in the ranking used only 53,405 cores, and delivered a theoretical peak performance of 1.11 petaflops, according to the Top 500 website.
Raijin cost AUD $50 million to build and will cost AUD $12 million a year to run, according to a story in the Australian Broadcasting Corporation. The cluster, which was named after the Japanese god of rain and thunder, will be used extensively with atmospheric and ocean modeling, in addition to weather forecasting and water resource management.
"You could say that we in the climate science community have a need for speed," the head of the Bureau of Meteorology Rob Vertessy told AFP. "The simple fact is that supercomputer capacity is a major determinant of our success in this field, but it's always been a struggle to secure access to it."
Andy Hogg, an associate professor at ANU and a researcher at NCI, says Raijin will provide a huge performance boost in his research into how currents flow in the Southern Ocean. When Hogg began his research 10 years ago, his HPC resource had just 8 processors. With Raijin, a single program can simultaneously utilize 3,000 to 4,000 cores, he says. "Our work is really underpinning the improvement in climate models that we'll see in the next five years," Hogg says in a video.
Raijin will also a boon to Australia's scientific community. Until now, Australian scientists have had to go elsewhere to gain access to the computational power required to run their models. With the powerful Raijin resource now available, those scientists can stay in the country.
The cluster is the highlight of the new NCI facility, which was also officially unveiled to the public today. In addition to Raijin, the facility will features a new 3,200-core cluster from Dell. That AUD $3.2 million Australian deal was announced last week.
Raijin is funded as part of Australia's infrastructure-building stimulus package in the wake of the global financial crisis, the AFP reports. NCI is funded by the ANU, the Commonwealth Scientific and Industrial Research Organization (CSIRO), the Australian Bureau of Meteorology, Geoscience Australia, and other universities through the Australian Research Council.
10/30/2013 | Cray, DDN, Mellanox, NetApp, ScaleMP, Supermicro, Xyratex | Creating data is easy… the challenge is getting it to the right place to make use of it. This paper discusses fresh solutions that can directly increase I/O efficiency, and the applications of these solutions to current, and new technology infrastructures.
10/01/2013 | IBM | A new trend is developing in the HPC space that is also affecting enterprise computing productivity with the arrival of “ultra-dense” hyper-scale servers.
Ken Claffey, SVP and General Manager at Xyratex, presents ClusterStor at the Vendor Showdown at ISC13 in Leipzig, Germany.
Join HPCwire Editor Nicole Hemsoth and Dr. David Bader from Georgia Tech as they take center stage on opening night at Atlanta's first Big Data Kick Off Week, filmed in front of a live audience. Nicole and David look at the evolution of HPC, today's big data challenges, discuss real world solutions, and reveal their predictions. Exactly what does the future holds for HPC?