[en] Published in summary form only

[en] Automatic parallelization of sequential programs will not achieve supercomputer performance in real applications. Instead, applications will have to be written with explicit parallelism. Machine-indepence of parallel programs is a precondition for wide acceptance of parallel computers. High Performance Fortran (HPF) will achieve machine-independence to an initially satisfactory degree, but another language revision can be expected. (orig./DG)

[en] Data communication between processing cores can waste running time and be optimized by performing computation whenever data communication occurs. A new method different from the MPI (message passing interface) has been developed and tested on 2 simulation codes. The running time for both codes has been reduced by a factor 1.55. (A.C.)

[en] Tensor algebra is a powerful tool for simulation codes and data analysis. Tensors are often partially empty and complex calculations should be implemented in a unique kernel for efficiency and memory managing. Code developers have to optimize mathematical expressions in which tensors appear, these tensors have various dense and empty dimensions which make them difficult to handle. TACO (acronym for tensor algebra compiler) is a compiler that allows an automated generation of parallel coding with optimized kernels and for an infinity of mathematical expressions. (A.C.)

[en] The current generation of UNIX farms at Fermilab are rapidly approaching the end of their useful life. The workstations were purchased during the years 1991-1992 and represented the most cost-effective computing available at that time. Acquisition of new workstations is being made to upgrade the UNIX farms for the purpose of providing large amounts of computing for reconstruction of data being collected at the 1996-1997 fixed-target run, as well as to provide simulation computing for CMS, the Auger project, accelerator calculations and other projects that require massive amounts of CPU. 4 refs., 1 fig., 2 tabs

[en] In this note, we present the results obtained with the CEA Benchmark on a CRAY Y-MP/832. The performances of the new machine are better than those of a Cray X-MP, specially regarding parallel computation

[en] We present a system which allows job submission over Internet with reliable data transfer using WWW graphical interface for job submission and monitoring and job management

[en] A parallel computational implementation of modern meshless methods is presented for explicit dynamic analysis. The procedures are demonstrated by application of the Reproducing Kernel Particle Method (RKPM). Aspects of a coarse grain parallel paradigm are detailed for a Lagrangian formulation using model partitioning. Integration points are uniquely defined on separate processors and particle definitions are duplicated, as necessary, so that all support particles for each point are defined locally on the corresponding processor. Several partitioning schemes are considered and a reduced graph-based procedure is presented. Partitioning issues are discussed and procedures to accommodate essential boundary conditions in parallel are presented. Explicit MPI message passing statements are used for all communications among partitions on different processors. The effectiveness of the procedure is demonstrated by highly deformable inelastic example problems

[en] The creation of a local area network (LAN) of interconnected computers provides an environment of multi computer processors that adds a new dimension to treatment planning. A LAN system provides the opportunity to have two or more computers working on the plan in parallel. With high speed interprocessor transfer, events such as the time consuming task of correcting several individual beams for contours and inhomogeneities can be performed simultaneously; thus, effectively creating a parallel multiprocessor treatment planning system

[en] Short communication