PVM for SCI Clusters

PVM-SCI is a complete implementation of the PVM (parallel virtual machine) message-passing environment for SCI clusters. It provides two complementary communication mechanisms: the conventional TCP and UDP protocols that are also used by the standard PVM for sending messages from one task to another via the PVM daemons, and a special-purpose protocol that utilizes the fast SCI network when the PvmRouteDirect option is set. In the latter case, no time-consuming software protocol is used, resulting in a significantly reduced communication latency and improved data throughput.

[1]  Hiroshi Tezuka,et al.  Pin-down cache: a virtual memory management technique for zero-copy communication , 1998, Proceedings of the First Merged International Parallel Processing Symposium and Symposium on Parallel and Distributed Processing.

[2]  Thorsten von Eicken,et al.  Incorporating Memory Management into User-Level Network Interfaces , 1997 .

[3]  Jack Dongarra,et al.  PVM: Parallel virtual machine: a users' guide and tutorial for networked parallel computing , 1995 .

[4]  Thomas M. Warschko,et al.  PULC: ParaStation User-Level Communication. Design and Overview , 1998, IPPS/SPDP Workshops.

[5]  Bernard Tourancheau,et al.  BIP: A New Protocol Designed for High Performance Networking on Myrinet , 1998, IPPS/SPDP Workshops.

[6]  Charles L. Seitz,et al.  Myrinet: A Gigabit-per-Second Local Area Network , 1995, IEEE Micro.

[7]  James Arthur Kohl,et al.  HARNESS: Heterogeneous Adaptable Reconfigurable NEtworked SystemS , 1998, Proceedings. The Seventh International Symposium on High Performance Distributed Computing (Cat. No.98TB100244).