When installing a new system (whether HPC, cloud, or just a bunch of servers, disks, etc), they must be housed. Certainly this can be without any specialist environment, especially if one is building a small test cluster; for example with half-a-dozen old but homogeneous systems, each connected with 100BASE-TX ethernet to a switch etc.
By default, EasyBuild will delete the build directory of an successful installation and will save failures of attempted installs for diagnostic purposes. In some cases however, one may want to save the build directory. This can be useful, for example, for diagnosis of *successful* builds. Another example is the installation of plugins for applications such as Paraview, which *require* access to the the successful buildir.
Recently, the University of Melbourne is ranked #1 in Australia and #33 in the world, according to the Times Higher Education World University Rankings of 2015-2016 . The rankings are based on a balanced metric between citations, industry income, international outlook, research, and teaching.
There is a wonderful Spanish idiom, "Cada loco con su tema" which is sometimes massacred as the English idiom "To each their own". In Spanish of course it is more accurately transliterated as "Each madman with their topic" which in familiar conversation means the same, has a slightly different and is a more illustrative angle on the subject. With the in mind, which path to insanity does one take with R libraries and EasyBuild? A similar question can also be raised with other languages that have extensions, e.g., Python and Perl.
The visit to Center for Scientific Computing (CSC) was carried out on October 14, 2016. Based at the Riedberg campus of the Goethe University Frankfurt the centre currently operates two Linux-based computer clusters FUCHS, and LOEWE-CSC. FUCHS has 14 air-cooled and 5 water-cooled racks using AMD Opteron (Istanbul and Magny-Cours) with 39956 cores total, mixed 4X DDR-QDR InfiniBand fabric and, a parallel scratch file system with an aggregated bandwidth of 6 GB/s and a capacity of 600 TB. 41 TFlops peak perfomance.
Following the acceptance of the University of Melbourne's paper to the OpenStack Barcelona summit in late October for their new HPC-cloud hybrid system, Spartan, an opportunity presented itself to visit other HPC centres and review their architecture and training programmes.
For about two weeks prior and a week after presenting at the OpenStack Summit in Barcelona I had the opportunity to visit several of Europe's major high performance computing facilities, giving each a bit of a standard pitch for the HPC-Cloud hybrid system we had developed at the University of Melbourne.
The somewhat nebulous term "supercomputer" has a long history. Although first coined in the 1920s to refer to IBMs tabulators, in electronic computing the most important initial contribution was the CDC6600 in the 1960s, due to its advanced performance over competitors. Over time major technological advancements included vector processing, cluster architecture, massive processors counts, GPGPU technologies, multidimensional torus architectures for interconnect.
COBOL is a business-orientated programming language that has been in use since 1959, making it one of the world's oldest programming languages.
Despite being much criticised (and for good reasons) it is still a major programming language in the financial sector, although there are a declining number of experienced programmers.
Thursday July 30th, at the Gryphon Gallery at the University of Melbourne, was the official launch of the 'Spartan' high-performance computing and cloud hybrid. Speakers at the launch included Dr Stephen Giugni, Director, Research Platform Services., Prof Margaret Sheil, Acting Vice Chancellor of the University of Melbourne., Professor Richard Sinnott, Director, eResearch and Professor of Applied Computing Systems., Mr Bernard Meade, Head of Research Compute Services, Research Platform Services, and yours truly, in my role as HPC Support Engineer, Research Platform Services.
As I argued in my presentation, the great advantage of Spartan is that it is designed around what users need. Based on research from the previous general compute resource, Edward, most people wanted to submit lots of jobs with a relatively small core count and memory footprint with data parallel approaches, but some really needed a large core counts with a fast interconnect. Putting the two types of users of the same system was not ideal. Also, engineers tend to want performance from a system, whereas managers want flexibility. Spartan provides both through its partitioning system. I am convinced that this will be architecture of future research computing.
Spartan's launch has received extensive media coverage, including high ranking sites such as HPC Wire, Gizmodo, and Delimiter. In addition to the aforementioned speakers, particular thanks must also be given to Linh Vu, Daniel Tosello, and Chris Samuel for their engineering excellence in helping put together the system, and to Greg Sauter for his project management (and for his photography). Welcome to Spartan!