Visit additional Tabor Communication Publications
July 26, 2012
While governments in much of the rest of the world are wringing their hands over stagnant or shrinking R&D budgets, Australia is buying up HPC machinery like there is no tomorrow. Just this week, Cray, IBM, and SGI announced supercomputing deals that would send the vendors' latest and greatest HPC equipment Down Under. In this case, the three systems are headed to various research facilities in New South Wales and Western Australia.
The biggest new Aussie system, by far, is the upcoming Cray "Cascade" supercomputer destined for the Pawsey Centre in Perth, Western Australia. That system will also include an unspecified amount of Sonexion storage, Cray's new Lustre-based line that they added to their portfolio last year. According to Cray, the total deal, including products and services, will amount to over $21 million.
Cascade is Cray's next-generation supercomputer, which is due out in 2013. It will be based on the upcoming Aries interconnect and support Intel processors – Xeon CPUs as well as the new Xeon Phi (MIC) manycore coprocessors. The Pawsey machine will incorporate both chips at least in the second phase of its deployment in 2014.
That system is expected to deliver 1.2 petaflops, which would make it about 40 percent more powerful than "Avoca," Australia's current top supercomputer. According to the Pawsey announcement, the machine will use "a combination of Intel Ivy Bridge, Haswell and MIC processors, although the precise configuration is still to be determined." The first iteration of the system, scheduled to be installed in 2013, will top out at 300 teraflops, which likely means no coprocessor acceleration.
Astronomers will use the Pawsey Cascade system to analyze data collected by the Australian Square Kilometer Array Pathfinder (ASKAP) and Murchison Widefield Array (MWA) radio telescopes. Besides sifting through astronomical data, the system will also be available to support research projects in other areas, including geosciences, nanotechnology and biosciences.
The Pawsey Supercomputing Centre was set up three years ago to support the SKA (Square Kilometre Array) radio telescope project, as well as a number of other scientific programs. The center is owned by Australia's major public research agency, the Commonwealth Scientific and Industrial Research Organisation (CSIRO), and is managed by iVEC, a joint supercomputing consortium that includes CSIRO, the Western Australia government, and a number of universities. Back in 2009, the feds gave $80 million to iVEC to set up the facility, which is now known as the Pawsey Centre. A big chunk of that funding is now going to the Cascade deployment.
Last September, iVEC acquired an SGI cluster, known as Fornax, also intended for radio astronomy support and general science duty. At 96 nodes, that system is much less powerful than the upcoming Cascade system. Each Fornax node contains two 6-core Intel Xeon Westmere CPUs and an NVIDIA Tesla GPU. Although the Fornax procurement took place last fall, the system apparently has yet to be deployed.
The machines currently up and running at Pawsey include HP and SGI clusters. The HP machine is an 87 teraflop ProLiant Blade system with 9600 cores of Westmere-generation CPUs, stitched together with QDR InfiniBand. It sports half a petabyte of external storage. The SGI Altix 3700 Bx2 is an even smaller system made up of 192 CPUs, 366 GB of memory, and 12 TB of disk. At 1.2 petaflops, the upcoming Cascade system will increase Pawsey's number crunching capacity by a factor of 10.
Radio telescope research has apparently become a growth industry Down Under. The second Australian supercomputer procurement announced this week is also slated to support the Murchison Widefield Array radio telescopes mentioned above. This machine is an IBM iDataPlex cluster and according to the press release will be used to "convert the radio waves into wide-field images of the sky that are unprecedented in clarity and detail."
The iDataPlex machine is a dx360 M3 rackmount cluster, build for density and energy efficiency. Each node houses two Xeon Westmere-EP CPUs and up to 192 GB of RAM. The specific size and configuration for the Murchison machine was not specified, but according to the announcement the cluster will be capable of crunching about 50 terabytes of data per day at a rate of up to 8 gigabytes per second.
In this case, the supercomputer will be housed on-site at Murchison Radio Observatory, which is located about 700 km north of Perth in the Australian Outback. That will give the system quick access to the data collected by the 4,096 radio antennas at the facility, enabling the images to be processed in real-time. And even though the system is being deployed in Australia, the Murchison radio telescope work is supported by an international consortium, which also includes New Zealand, India, and the US.
The third new Aussie machine is an SGI cluster, and is headed to Sydney, on the other side of the continent. The system was procured by Intersect, a local non-profit group that provides IT services, including HPC, to public and private researchers in New South Wales.
At 33 teraflops, the SGI system is medium-sized by today's standards. It consists of 100 nodes -- 88 small compute nodes, 10 large ones, and 2 admin/system console nodes. All are of the dual-socket Sandy Bridge (Xeon E5-2600) persuasion, with 256 GB of memory on the large nodes and 64 GB on the small ones. Everything is hooked together with QDR InfiniBand. The procurement also includes 101 TB of storage.
According to the announcement, the new cluster will support research in areas such as quantum chemistry, computational chemistry, chemical engineering, climate science, mechanical engineering, bioinformatics and physics. In this case, no mention was made of radio telescope work, most of which is located in Western Australia.
It was probably just a coincidence that these three HPC deals were announced a few days apart. Australia doesn't have the wherewithal to ramp up its HPC capability and become a world leader in supercomputing anytime soon. With an R&D budget about 1/20th the size of the US, the country can only do so much.
However compared to much of the world, Australia is in an enviable position right now. The economy there is enjoying solid and sustained growth, with low unemployment, low interest rates, manageable debt, and a nice investment pipeline. Even a slowing economy in China, a recessionary environment in Europe, and an uneven recovery in the US are unlikely faze the Australians very much. If there was ever a time to gain some ground in supercomputing, this would be it.
Jun 18, 2013 |
The world's largest supercomputers, like Tianhe-2, are great at traditional, compute-intensive HPC workloads, such as simulating atomic decay or modeling tornados. But data-intensive applications--such as mining big data sets for connections--is a different sort of workload, and runs best on a different sort of computer.
Jun 18, 2013 |
Researchers are finding innovative uses for Gordon, the 285 teraflop supercomputer housed at the San Diego Supercomputer Center (SDSC) that has a unique Flash-based storage system. Since going online, researchers have put the incredibly fast I/O to use on a wide variety of workloads, ranging from chemistry to political science.
Jun 17, 2013 |
The advent of low-power mobile processors and cloud delivery models is changing the economics of computing. But just as an economy car is good at different things than a full size truck, an HPC workload still has certain computing demands that neither the fastest smartphone nor the most elastic cloud cluster can fulfill.
Jun 14, 2013 |
For all the progress we've made in IT over the last 50 years, there's one area of life that has steadfastly eluded the grasp of computers: understanding human language. Now, researchers at the Texas Advanced Computing Center (TACC) are utilizing a Hadoop cluster on its Longhorn supercomputer to move the state of the art of language processing a little bit further.
Jun 13, 2013 |
Titan, the Cray XK7 at the Oak Ridge National Lab that debuted last fall as the fastest supercomputer in the world with 17.59 petaflops of sustained computing power, will rely on its previous LINPACK test for the upcoming edition of the Top 500 list.
05/10/2013 | Cleversafe, Cray, DDN, NetApp, & Panasas | From Wall Street to Hollywood, drug discovery to homeland security, companies and organizations of all sizes and stripes are coming face to face with the challenges – and opportunities – afforded by Big Data. Before anyone can utilize these extraordinary data repositories, however, they must first harness and manage their data stores, and do so utilizing technologies that underscore affordability, security, and scalability.
04/15/2013 | Bull | “50% of HPC users say their largest jobs scale to 120 cores or less.” How about yours? Are your codes ready to take advantage of today’s and tomorrow’s ultra-parallel HPC systems? Download this White Paper by Analysts Intersect360 Research to see what Bull and Intel’s Center for Excellence in Parallel Programming can do for your codes.
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?
Join our webinar to learn how IT managers can migrate to a more resilient, flexible and scalable solution that grows with the data center. Mellanox VMS is future-proof, efficient and brings significant CAPEX and OPEX savings. The VMS is available today.