ABSTRACT
Parallel programming is developing very fast and it is already part of several aspects of our everyday life. Furthermore, the fall of prices for hardware along with an increase in terms of reliability and performance, the wider and wider availability of free software and the wide usage and need for parallel computing and processing have created the natural environment for a development (an almost spontaneous evolution) that has lead to Beowulf, a technology to create a parallel supercomputer out of a cluster of Linux boxes. Nowadays it is possible to create parallel computers effectively at a reasonable price by using off-the-shelf technology. In fact it is possible to build a parallel computer at home by interconnecting components that can be found at an electronic shop. Moreover, we have to deal with the application rather than the specific hardware which carries it out. In fact, universal, general purpose parallel machines simply do not exist. The application must be designed before the cluster architecture. Berserkr represents the solution for this set of problems: it aims to be a tool to test and compare different possible implementations (both in hardware and in software) on several different architectures by building the hardware structure just in the virtual world and not physically.
- M. Ben-Ari and Y. B.-D. Kolikant. Thinking parallel: the process of learning concurrency. ACM ITiCSE 1999 6/99, Cracow, Poland, 1999. Google ScholarDigital Library
- E. Carniani and R. Davoli. The netwire emulator: a tool for teaching and understanding networks. ACM SIGCSE Bulletin, 33(3):153--152, June 2001. Google ScholarDigital Library
- A. L. Fisher and T. Gross. Teaching the programming of parallel computers.ACM SIGCSE Bulletin, 23(1):102--107, March 1991. Google ScholarDigital Library
- D. D. C. Hyde. A parallel processing course for undergraduates. Twentieth SIGCSE Technical Symposium, pages 170--173, February 1989. Google ScholarDigital Library
- L. Jin and L. Yang. A laboratory for teaching parallel computing on parallel structures. In Proceedings of the 26th SIGCSE Technical Symposium on Computer Science Education, Nashville, Tennessee, USA, pages 71--75, March 1995. Google ScholarDigital Library
- D. J. John. Nsf supported projects: Parallel computation as an integrated component in the undergraduate curriculum in computer science. In Proceedings of the 25th SIGCSE Technical Symposium on Computer Science Education, 26:357--361, 1994. Google ScholarDigital Library
- U. of Roma3. netkit. http://www.dia.uniroma3.it/, 2003.Google Scholar
- J. Radajewsky and D. Eadline. Beowulf how to, version 1.1.1 22. http://www.sci.usq.edu.au/staff/jacek/beowulf, 1998.Google Scholar
- M. Rechenburg. User mode openmosix. http://www.mosixview.com/umopenmosix, 2003.Google Scholar
- F. G. Tinetti. Performance of scientific processing in networks of workstations: Matrix multiplication example. http://journal.info.unlp.edu.ar/journal/journal4/papers/pap7.pdf, 2001.Google Scholar
- W. E. Toll. Decison point in the introduction of parallel processing into the undergradaute curriculum. In Proceedings of the 26th SIGCSE Technical Symposium on Computer Science Education, Nashville, Tennessee, 27:136--140, March 1998. Google ScholarDigital Library
- UML. User mode linux. http://user-mode-linux.sourceforge.net/, 2003.Google Scholar
Index Terms
- Berserkr: a virtual beowulf cluster for fast prototyping and teaching
Recommendations
MPPs and clusters for scalable computing
ISPAN '96: Proceedings of the 1996 International Symposium on Parallel Architectures, Algorithms and NetworksThis article assess the state-of-the-art technology in massively parallel processors (MPPs) and clusters of workstations (COWs) for scalable parallel computing. We evaluate the IBM SP2, the Intel Paragon, the Cray T3D/T3E, and the ASCI TeraFLOPS system ...
Performance evaluation of supercomputers using HPCC and IMB Benchmarks
The HPC Challenge (HPCC) Benchmark suite and the Intel MPI Benchmark (IMB) are used to compare and evaluate the combined performance of processor, memory subsystem and interconnect fabric of five leading supercomputers-SGI Altix BX2, Cray X1, Cray ...
Aizu supercomputer: a massively parallel system for virtual reality problems
PAS '95: Proceedings of the First Aizu International Symposium on Parallel Algorithms/Architecture SynthesisThe Aizu supercomputer is a massively parallel system suited to the solution of virtual reality problems and the support of multimedia applications. It employs a highly parallel MIMD architecture using a conflict-free internetwork system. The scalable ...
Comments