A wide variety of physical phenomena arising within many scientific disciplines can be described by systems of coupled partial differential equations (PDEs). The numerical approximation of these PDEs often involves the solution of a system of algebraic equations (possibly nonlinear) which are typically large, sparse and nonsymmetric. The increasing computational demands required by the solution of such complex scientific applications has motivated the current direction toward large-scale parallel computers. We, therefore, consider solution techniques of representative systems of equations on large scale MIMD machines. Our primary emphasis in this study is the evaluation of iterative methods for the solution of nonsymmetric systems. In particular, we discuss two Krylov subspace methods, the conjugate gradient squared algorithm (CGS) and the generalized minimum residual method (GMRES), along with the multigrid algorithm (MG) on massively parallel MIMD architectures. The focus of this evaluation considers the performance of various algorithm and implementation variations over a broad selection of problems using a parallel machine.

The purpose of this writeup is to supplement the publication, Massively Parallel Fokker-Planck Calculations,'' which appeared in the Proceedings of the Fifth Distributed Memory Computing Conference. Results obtained subsequent to that presentation are reported. This work is not self-contained; the reader should refer to that publication.

The Notre Dame Radiation Laboratory is a facility of the Department of Energy operated for DOE by the University of Notre Dame. The majority of the programs within the Laboratory are supported by the Office of Basic Energy Sciences of DOE and, unless otherwise noted in the following, support can be attributed to the Office of Basic Energy Sciences. Certain additional programs, so indicated, are supported by the Office of Standard Reference Data of the National Institute of Standards and Technology. Research programs conducted at the laboratory are briefly described.

The Fokker-Planck package FPPAC had been converted to the Connection Machine 2 (CM2). For fine mesh cases the CM2 outperforms the Cray-2 when it comes to time-integrating the difference equations. For long Legendre expansions the CM2 is also faster at computing the Fokker-Planck coefficients. 3 refs.

To participate in the 1990 CORA Workshop at Kernforschungszentrum Karlsruhe (KfK) GmbH, Karlsruhe, FRG, on October 1--4, and to participate in detailed discussions on October 5 with the KfK CORA Boiling Water Reactor (BWR) experiments. The traveler attended the 1990 CORA Workshop at KfK, FRG. Participation included the presentation of a paper on work performed by the Boiling Water Reactor Core Melt Progression Phenomena Program at Oak Ridge National Laboratory (ORNL) on posttest analyses of CORA BWR experiments. The Statement of Work (November 1989) for the BWR Core Melt Progression Phenomena Program provides for pretest and posttest analyses of the BWR CORA experiments performed at KfK. Additionally, it is intended that ORNL personnel participate in the planning process for future CORA BWR experiments. For these purposes, meetings were held with KfK staff to discuss such topics as (1) experimental test schedule, (2) BWR test conduct, (3) perceived BWR experimental needs, and (4) KfK operational staff needs with respect to ORNL support. 19 refs.

A weekly publication, the Texas Register serves as the journal of state agency rulemaking for Texas. Information published in the Texas Register includes proposed, adopted, withdrawn and emergency rule actions, notices of state agency review of agency rules, governor's appointments, attorney general opinions, and miscellaneous documents such as requests for proposals. After adoption, these rulemaking actions are codified into the Texas Administrative Code.

Document issued by the Office of the Attorney General of Texas in Austin, Texas, providing an interpretation of Texas law. It provides the opinion of the Texas Attorney General, Jim Mattox, regarding a legal question submitted for clarification; Authority of a county clerk to collect a fee upon filing of an application for emergency detention or court-ordered services in cases involving mental competency or chemical dependency, andrelated questions (RQ-2004)

There is an OR module assembly stand in use at IB4. This design has been approved by safety, as presented by Mike Foley, and has been successfully used. Another one is needed at the D-zero assembly building, but some modifications need to be made. This report will show that the new modified design is at least as strong, if not stronger, than the older IB4 design in every aspect. Since the weight distribution of the OR modules on the sling is indeterminate, this report compares three cases of support for the entire assembly: the lowest two beams only, the lowest four beams only, and all six beams. In each of these cases, the new design is stronger than the old design in maximum allowable weight. The ability of the the cradle to support the weight is also shown. For all of the failure conditions except for two, the cradle is stronger than the beams that it supports. In the two excepted situations, the calculated limit of the cradle is less than the beams it supports. This is because no credit is taken for the sling and strongback, which in reality will relieve much of the horizontal load.