Network-based multicomputers: redefining high performance computing in the 1990s

One of the most significant advances in computer systems over the past decade is parallel processing. This is a direct consequence of close interactions between system design and VLSI technology. This integrated approach has allowed insights in computations, systems and applications to have fundamental influences on the design of hardware structures, and vice versa. The cooperation will be even more crucial as we enter the next decade when parallel computers are expected to be not only powerful in performance but also easy to use. A key point here is to use the power of VLSI to implement general and very high performance networks for large-scale multicomputer. In this paper we briefly review current work at Carnegie Mellon in this area, and give a taxonomy to show the general architectural trends. It is concluded that multicomputer based on flexible, extensible and efficient network backplanes will be a major thrust in high performance computing in the 1990s. The research was supported in part by Defense Advanced Research Projects Agency (DOD) monitored by the Space and Naval Warfare Systems Command under Contract N00039-87-C-0251, and in part by the Office of Naval Research under Contracts N00014-87-K-0385 and N00014-87-K-0533. The views and conclusions contained in this document are those of the author and should not be interpreted as representing the official policies, either expressed or implied, of the Defense Advanced Research Projects Agency, the Office of Naval Research or the U.S. government. Decennial Caltech Conference on VLSI Pasadena, California, 20-22 March 1989 Network-Based Multicomputer: Redefining High Performance Computing in the 1990s

[1]  H. T. Kung.  Deadlock avoidance for systolic communication , 1988, ISCA 1988.

[2]  Shekhar Y. Borkar,et al.  iWarp: an integrated solution to high-speed parallel computing , 1988, Proceedings. SUPERCOMPUTING '88.

[3]  H. T. Kung,et al.  The design of nectar: a network backplane for heterogeneous multicomputers , 1989, ASPLOS III.

[4]  Thomas R. Gross,et al.  Architecture and compiler tradeoffs for a long instruction wordprocessor , 1989, ASPLOS III.

[5]  Han Wang,et al.  Design and analysis of image processing algorithms for programmable VLSI array processors , 1988 .

[6]  Charles L. Seitz,et al.  Multicomputers: message-passing concurrent computers , 1988, Computer.

[7]  L. W. Tucker,et al.  Architecture and applications of the Connection Machine , 1988, Computer.

[8]  H. T. Kung,et al.  A Highly Reconfigurable Array Of Powerful Processors , 1988, Optics & Photonics.

[9]  Thomas R. Gross,et al.  Compilation for a high-performance systolic array , 1986, SIGPLAN '86.

[10]  Alain J. Martin,et al.  The architecture and programming of the Ametek series 2010 multicomputer , 1988, C3P.

[11]  William J. Dally,et al.  A VLSI Architecture for Concurrent Data Structures , 1987 .

[12]  H. T. Kung.  Deadlock avoidance for systolic communication , 1988, ISCA '88.

[13]  Monica Sin-Ling Lam,et al.  A Systolic Array Optimizing Compiler , 1989 .

[14]  Mark Homewood,et al.  The IMS T800 Transputer , 1987, IEEE Micro.

[15]  H. T. Kung.  Systolic communication , 1988, [1988] Proceedings. International Conference on Systolic Arrays.

[16]  Charles L. Seitz,et al.  The cosmic cube , 1985, CACM.

[17]  Leonard Kleinrock,et al.  Virtual Cut-Through: A New Computer Communication Switching Technique , 1979, Comput. Networks.

[18]  E. Chow,et al.  A high-speed message-driven communication architecture , 1988, ICS '88.

[19]  H. T. Kung,et al.  The Warp Computer: Architecture, Implementation, and Performance , 1987, IEEE Transactions on Computers.

[20]  Dirk Grunwald,et al.  Hyperswitch network for the hypercube computer , 1988, ISCA '88.