On dynamic parallel execution of large scientific codes /
allocate free processors to multiple tasks at their requests
| Main Author: | |
|---|---|
| Format: | Thesis eBook |
| Language: | English |
| Published: |
[Place of publication not identified] :
[publisher not identified] ;
1994.
|
| Subjects: | |
| Online Access: | Link to OAKTrust copy |
| Summary: | allocate free processors to multiple tasks at their requests analyzer can communicate to the compiler to insert special application codes can be effectively exploited. By inserting special tags in the code the parallelism Detection and utilization of the natural parallelism of a features the Dynamic Processor Allocator coprocessor can Fortran language are analyzed to study the looping behavior instructions for the dynamic allocation of processors at run- massively parallel processing (MPP) computing system. In of these codes to determine loops that merit parallelization. program is the key to exploiting the computing power of a simulation results show that the proposed scheme utilizes the system resources only when necessary, and the parallelism of this thesis some large scientific codes written in the through a dynamic processor allocation protocol. The time of an MPP system. The target MPP architecture which |
|---|---|
| Item Description: | "Major subject: Electrical Engineering". Vita. |
| Physical Description: | vi, 58 leaves : illustrations ; 28 cm. Also available online. |
| Bibliography: | Includes bibliographical references. |