No abstract

SUMMARYIn the near future, interconnection networks of massively parallel computer systems will connect more than a hundred thousands of computing nodes. The performance evaluation of the interconnection networks can provide real insights to help the development of efficient communication library. Hence, to evaluate the performance of such interconnection networks, simulation tools capable of modeling the networks with sufficient details, supporting a user-friendly interface to describe communication patterns, providing the users with enough performance information, completing simulations within a reasonable time, are a real necessity. This paper introduces a novel interconnection network simulator NSIM, for the evaluation of the performance of extreme-scale interconnection networks. The simulator implements a simplified simulation model so as to run faster without any loss of accuracy. Unlike the existing simulators, NSIM is built on the execution-driven simulation approach. The simulator also provides a MPI-compatible programming interface. Thus, the simulator can emulate parallel program execution and correctly simulate point-to-point and collective communications that are dynamically changed by network congestion. The experimental results in this paper showed sufficient accuracy of this simulator by comparing the simulator and the real machine. We also confirmed that the simulator is capable of evaluating ultra large-scale interconnection networks, consumes smaller memory area, and runs faster than the existing simulator. This paper also introduces a simulation service built on a cloud environment. Without installing NSIM, users can simulate interconnection networks with various configurations by using a web browser.

Studies in the field of computer science cover aspects of science that attach importance to analysis and aspects of engineering that attach importance to synthesis. Therefore, the ability to manage synthesis, represented by system development and/or its design, is required in addition to analysis in the field. On the other hand, essential course subjects in computer science education such as logic circuits, computer architecture, operating system, and compilers have to be closely related with one another; nevertheless, they tend to be isolated from one another because of their great sophistication and complication. Thus, consistent education in computer science becomes difficult. Moreover, experiments, not virtual experiences, which involve the pleasure and emotion of creation are necessary in hardware education. Accordingly, since the beginning of the 1990s, we have developed an educational microprocessor called KITE using reconfigurable LSI, together with teaching materials that allow students to learn actively, considering the essence of synthesis as related to comprehensive understanding of lectures and practical use. These teaching materials have been made available to the public, and have been introduced into more than 30 educational facilities such as universities and companies. They are well designed for various stages of computer education, varying from introductory education to design education, and system software education, providing an effective education, with experiments in addition to lectures, and a consistent grounding in computer science. © 2002 Wiley Periodicals, Inc. Syst Comp Jpn, 33(8): 64–74, 2002; Published online in Wiley InterScience (http://www.interscience.wiley.com). DOI 10.1002/scj.10038

No abstract

SUMMARYThis paper proposes the recursive diagonal torus (RDT), which is the interconnection network for the massively parallel computer with 216 or more processors. RDT can realize a small diameter (12 for 16 Knode) with a small number of links compared to the hypercube by adding links recursively along the diagonal direction of the two-dimensional torus. The packet is routed by a simple algorithm calledvector routing. RDT not only includes a torus, but also easily implements the emulation of the hypercube or tree as well as the packet broadcast. Considering a typical packet transfer method, the random transfer performance is evaluated. A satisfactory result is obtained compared to the 2-D or 3-D torus, although somewhat inferior to the hypercube with a large number of links.