Browse Prior Art Database

SUBMICRON SYSTEMS ARCHITECTURE ' Semiannual Technical Report

IP.com Disclosure Number: IPCOM000127951D
Original Publication Date: 1986-Dec-31
Included in the Prior Art Database: 2005-Sep-14

Publishing Venue

Software Patent Institute

Related People

Charles L Seitz: AUTHOR [+3]

Abstract

The central theme of this research is the architecture and design of VLSI systems appropriate to a microcircuit technology scaled to submicron feature sizes, and includes related efforts in concurrent computation and VLSI design. Additional background information can be found in previous semiannual tech- " nical reports [5052:TR:82], [5078:TR:83], [5103:TR:83], (5122:TR:84), [5160:TR:84], (5178:TR:851, (5202:TR:851. 1.3 Highlights Some highlights of the previous 6 months are: * Working Mosaic A elements (2.2 and 4.1)!!! * Torus Routing Chips work on first silicon (4.2). * Mosaic C - an SCMOS Mosaic element - being designed (2.2.2 and 4.3). * Compilation method for designing self-timed circuits (4.5). * 128-node Intel iPSC in routine operation and available for use via the ARPAnet (2.1.3). * New results in parallel execution of logic programs (3.1). * A new flow framework for concurrent programming (3.3) * Event-driven simulations chew up endless cycles on the cosmic cubes (3.5). * A surprise about network topologies (3.6).

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 3% of the total text.

Page 1 of 42

THIS DOCUMENT IS AN APPROXIMATE REPRESENTATION OF THE ORIGINAL.

SUBMICRON SYSTEMS ARCHITECTURE ' Semiannual Technical Report

Department of Computer Science California Institute of Technology

5220:TR:86 March 1986 Reporting Period: 16 September 1985 to 15 March Principal Investigator: Charles L Seitz

Faculty Investigators: James T Kajiya Alain J Martin Robert J McEliece Martin Rem Charles L Seitz Sponsored by the Defense Advanced Research Projects Agency ARPA Order Number 3'771

Monitored by the Office of Naval Research Contract Number N00014-79 SUBMICRON SYSTEMS ARCHITECTURE Department of Computer Science California Institute of Technology

1. Overview and Summary

1.1 Scope of this Report

This document is a summary of the research activities and results for the six month period 16 September 1985 to 15 March 1986 under the Defense Advanced Research Project Agency (DARPA) Submicron Systems Architecture Project. Tech-nical reports covering parts of the project in detail are listed following these sum-maries, and can be ordered from the Caltech Computer Science Library.

1.2 Objectives

The central theme of this research is the architecture and design of VLSI systems appropriate to a microcircuit technology scaled to submicron feature sizes, and includes related efforts in concurrent computation and VLSI design. Additional background information can be found in previous semiannual tech- " nical reports [5052:TR:82], [5078:TR:83], [5103:TR:83], (5122:TR:84), [5160:TR:84], (5178:TR:851, (5202:TR:851. 1.3 Highlights Some highlights of the previous 6 months are: * Working Mosaic A elements (2.2 and 4.1)!!! * Torus Routing Chips work on first silicon (4.2). * Mosaic C - an SCMOS Mosaic element - being designed (2.2.2 and
4.3). * Compilation method for designing self-timed circuits (4.5). * 128-node Intel iPSC in routine operation and available for use via the ARPAnet (2.1.3). * New results in parallel execution of logic programs (3.1). * A new flow framework for concurrent programming (3.3) * Event-driven simulations chew up endless cycles on the cosmic cubes (3.5). * A surprise about network topologies (3.6).

2. Architectural Experiments

California Institute of Technology Page 1 Dec 31, 1986

Page 2 of 42

SUBMICRON SYSTEMS ARCHITECTURE ' Semiannual Technical Report

2.1 Cosmic Cube Project

W C Athas, Michael Lichter, Wen-King Su, Chuck Seitz

The cosmic cubes are working reliably, and researchers are using them for ap-plication programming projects and for event-driven simulations of other message-passing architectures. Usage has been moderately heavy, and there have been very few problems reported.

2.1.1 Hardware Status

The 6-cube interface was moved from sol.caltech.edu, our main SUN file-server, a SUN-2/170, to cares, a SUN-2/120, in order to lighten the load on the already overloaded file server. The 3- cube interface was moved from cares to venus, another SUN-2/120.

Neither the 6-cube nor the 3-cube has failed in this 6-month period. The c...