Initial Experiences from TUPAC Supercomputer
High Performance Computing centers boost the development of a wide range of disciplines in Science and Engineering. The installation of public shared facilities in a country is an effort that should be carefully used and must be as open as possible to strengthen the impact of these resources.
We describe the current status and characteristics of TUPAC supercomputer, which is hosted by a CONICET research institute. Unlike other experiences in Argentina, TUPAC is focused on supporting external scientific and technological communities. In spite of having a reduced operations staff, this machine provides computational resources and supports more than 200 external research projects.
In this work we describe the supercomputer setup, tools and policies implemented to reach the level of efficiency needed to support this amount of projects. We also characterize jobs, users and projects with special emphasis on industrial applications and large computational research initiatives hosted in TUPAC.
KeywordsHigh performance computing facilities Operations Monitoring
- 1.Centro de simulación computacional p/aplic tecnológicas. http://www.csc-conicet.gob.ar/. Accessed 2 Dec 2017
- 2.INVAP S.E.: Company devoted to the design and construction of complex technological systems. http://www.invap.com.ar/. Accessed 2 Dec 2017
- 3.Redmine: A flexible project management web application written using ruby on rails framework. http://www.redmine.org. Accessed 2 Dec 2017
- 4.ANSYS: Fluent software (2015). http://www.ansys.com/Products/Fluids/ANSYS-Fluent. Accessed 2 Dec 2017
- 5.Barth, W.: Nagios: System and Network Monitoring. No Starch Press, San Francisco (2006)Google Scholar
- 6.Hafner, J., Kresse, G., Vogtenhuber, D., Marsman, M.: Vienna Ab initio simulation package (2017). http://www.vasp.at/. Accessed 2 Dec 2017
- 7.Laizet, S., Lamballais, E.: High-order compact schemes for incompressible flows: a simple and efficient method with quasi-spectral accuracy. J. Comput. Phys. 228(16), 5989–6015 (2009). http://www.sciencedirect.com/science/article/pii/S0021999109002587 MathSciNetCrossRefzbMATHGoogle Scholar
- 9.Massie, M., Li, B., Nicholes, B., Vuksan, V., Alexander, R., Buchbinder, J., Costa, F., Dean, A., Josephsen, D., Phaal, P., Pocock, D.: Monitoring with Ganglia, 1st edn. O’Reilly Media Inc., Sebastopol (2012)Google Scholar
- 10.Pascual, J.A., Navaridas, J., Miguel-Alonso, J.: Effects of topology-aware allocation policies on scheduling performance. In: Frachtenberg, E., Schwiegelshohn, U. (eds.) JSSPP 2009. LNCS, vol. 5798, pp. 138–156. Springer, Heidelberg (2009). https://doi.org/10.1007/978-3-642-04633-9_8 CrossRefGoogle Scholar
- 13.Springel, V., Yoshida, N., White, S.D.: GADGET: a code for collisionless and gasdynamical cosmological simulations. New Astron. 6(2), 79–117 (2001). http://www.sciencedirect.com/science/article/pii/S1384107601000422 CrossRefGoogle Scholar
- 15.The OpenFOAM Foundation: Open source software for computational fluid dynamics (CFD). http://www.openfoam.org. Accessed 2 Dec 2017