Load Balancing Hpf Programs by Migrating Virtual Processors Ecole Normale Supérieure De Lyon Load Balancing Hpf Programs by Migrating Virtual Processors Load Balancing Hpf Programs by Migrating Virtual Processors
نویسنده
چکیده
This paper explores the integration of load balancing features in the data parallel lan guage HPF targeting semi regular applications We show that the HPF virtual pro cessors are good candidates to be the unit of migration Then we compare possible implementations and show that threads provide a good tradeo between e ciency and ease of implementation We nally describe a preliminary implementation The ex perimental results obtained with the Gaussian elimination with partial pivoting are promising
منابع مشابه
Load Balancing HPF programs by Migrating Virtual Processors
This paper explores the integration of load balancing features in the data parallel language HPF targeting semi-regular applications. We show that the HPF virtual processors are good candidates to be the unit of migration. Then, we compare 3 possible implementations and show that threads provide a good tradeoo between eeciency and ease of implementation. We nally describe a preliminary implemen...
متن کاملVisualization of Hpf Data Mappings and of Their Communication Cost
HPF-Builder graphical environment provides an interactive and visual solution to edit and visualize HPF data mapping directives. It frees the HPF programmers of all the syntactic constraints. General and detailled visualizations give complete information about data distribution along the grids of processors. Compare several mappings implies to evaluate some statistics about load distribution an...
متن کاملIterative Dynamic Load Balancing in Multicomputers
Dynamic load balancing in multicomputers can improve the utilization of processors and the eeciency of parallel computations through migrating workload across processors at runtime. We present a survey and critique of dynamic load balancing strategies that are iterative: workload migration is carried out through transferring processes across nearest neighbor processors. Iterative strategies hav...
متن کاملBlock Cyclic Array Redistribution Ecole Normale Supérieure De Lyon Block Cyclic Array Redistribution
Implementing linear algebra kernels on distributed memory parallel computers raises the problem of data distribution of matrices and vectors among the processors Block cyclic distribution seems to suit well for most algorithms But one has to choose a good compromise for the size of the blocks to achieve a good e ciency and a good load balancing This choice heavily depends on each operation so i...
متن کاملA user-Level process package for PVM
This paper describes an approach to supporting efcient processor virtualization and dynamic load balancing for message-based, parallel programs. Specifically, a user-level process package (UPVM) for SPMD-style PVM applications is presented. UPVM supports light-weight virtual processors that are transparently and independently migratable. It also implements a source-code compatible PVM interface...
متن کامل