MPI-2单边通信存在很高的通信开销以及对通信进程中远程进程的依赖。为此提出了在InfiniBand体系结构上的高性能MPI-2单边通信设计方法。其中,MPI-2单边通信操作,比如MPI_Put,MPI_Get以及MPI_Accumulate将对应于InfiniBand远程直接内存...MPI-2单边通信存在很高的通信开销以及对通信进程中远程进程的依赖。为此提出了在InfiniBand体系结构上的高性能MPI-2单边通信设计方法。其中,MPI-2单边通信操作,比如MPI_Put,MPI_Get以及MPI_Accumulate将对应于InfiniBand远程直接内存访问(Remote Direct MemoryAccess,RDMA)操作。设计是基于MPICH2的在InfiniBand上的应用,可以很好地实现通信和计算的重叠处理。展开更多
By begiwhng with the Cannon algorithm[1] and the double- direct ion dat amoving algoritlun[2] for processors arranged as a 2-D square mesh, we improvethese two algorithms to general 2-D mesh in this paper. We also con...By begiwhng with the Cannon algorithm[1] and the double- direct ion dat amoving algoritlun[2] for processors arranged as a 2-D square mesh, we improvethese two algorithms to general 2-D mesh in this paper. We also consider theapplication of our algorithms in those operations which are similar to matrix multiplication. Using MPI parallel programming environment, we have obtained satisfactory performance on Dawning-1000.展开更多
This paper describes a parallel computing platform using the existing facilities for the digital watershed model. In this paper, distributed multi-layered structure is applied to the computer cluster system, and the M...This paper describes a parallel computing platform using the existing facilities for the digital watershed model. In this paper, distributed multi-layered structure is applied to the computer cluster system, and the MPI-2 is adopted as a mature parallel programming standard. An agent is introduced which makes it possible to be multi-level fault-tolerant in software development. The communication protocol based on checkpointing and rollback recovery mechanism can realize the transaction reprocessing. Compared with conventional platform, the new system is able to make better use of the computing resource. Experimental results show the speedup ratio of the platform is almost 4 times as that of the conventional one, which demonstrates the high efficiency and good performance of the new approach.展开更多
文摘MPI-2单边通信存在很高的通信开销以及对通信进程中远程进程的依赖。为此提出了在InfiniBand体系结构上的高性能MPI-2单边通信设计方法。其中,MPI-2单边通信操作,比如MPI_Put,MPI_Get以及MPI_Accumulate将对应于InfiniBand远程直接内存访问(Remote Direct MemoryAccess,RDMA)操作。设计是基于MPICH2的在InfiniBand上的应用,可以很好地实现通信和计算的重叠处理。
文摘By begiwhng with the Cannon algorithm[1] and the double- direct ion dat amoving algoritlun[2] for processors arranged as a 2-D square mesh, we improvethese two algorithms to general 2-D mesh in this paper. We also consider theapplication of our algorithms in those operations which are similar to matrix multiplication. Using MPI parallel programming environment, we have obtained satisfactory performance on Dawning-1000.
基金the Creative Research Team Foundation of the National Natural Science Foundation of China (No. 50221903)
文摘This paper describes a parallel computing platform using the existing facilities for the digital watershed model. In this paper, distributed multi-layered structure is applied to the computer cluster system, and the MPI-2 is adopted as a mature parallel programming standard. An agent is introduced which makes it possible to be multi-level fault-tolerant in software development. The communication protocol based on checkpointing and rollback recovery mechanism can realize the transaction reprocessing. Compared with conventional platform, the new system is able to make better use of the computing resource. Experimental results show the speedup ratio of the platform is almost 4 times as that of the conventional one, which demonstrates the high efficiency and good performance of the new approach.
