ftp://pong.gsfc.nasa.gov/www_ftp/hpcc/dynamics/dycore_mp.V2_1.src.tar.Z Atmospheric Dynamics Benchmark Source Codes <author> <version> <abstract> The dynamics benchmark computes the equations of motion for a grid-point atmospheric global climate model. This benchmark contains nearest-neighbor communication, data redistribution and static load-balancing. <description><url>http://pong.gsfc.nasa.gov/hpcc/</url> <reference><url></url> <contact>Thomas M. Head / tom@elsa.gsfc.nasa.gov <keywords>application program <category>application <publication-date> <environment>The source code, written in fortran 77, has been designed to execute efficiently on sequential, vector and mpp architectures with minimal updating to the source code. The current version of the dynamics benchmark has been successfully executed on Sun, DEC Alpha and SGI Power Challange workstations, the Cray C90 vector processor, the CRAY T3D, IBM SP2 and Intel Paragon MPP's, as well as PVM implementations on networked Sun and Alpha workstations. <method> <application>atmospheric dynamics; global climate modeling <comments> </urc> <urc> <url> <title>Atmospheric Radiation Benchmark Source Codes <author> <version> <abstract> The radiation benchmarks compute long and short wave radiation values for a grid-point atmospheric global climate model. These routines are perfectly parallel codes, requiring communication only to set up the problem and gather checksum results. There is also a dynamic load-balancing version of the shortwave radiation code that insures an even distribution of work among processing nodes at the expense of increased communications. <description><url>http://pong.gsfc.nasa.gov/hpcc/</url> <reference><url></url> <contact> <keywords>application program <category>application <publication-date> <environment>The source code, written in fortran 77, has been designed to execute efficiently on sequential, vector and mpp architectures with minimal updating to the source code. The current version of the radiation benchmark has been successfully executed on Sun and DEC Alpha workstations, the Cray C90 vector processor, and Cray T3D, IBM SP2 and Intel Paragon MPP's. <method> <application>global climate model <comments> </urc>