We describe our efforts to scale message driven applica-tions to a large number of processors on an Alpha cluster interconnected by QsNet. The clustering technology QsNet has a network interface with a communication co-processor. The presence of the co-processor minimizes main processor participation in message passing. We show the advantages of the communication co-processor for message driven ap-plications. To scale fine-grained message driven appli-cations to a large number of processors we had to over-come several hardware, software and operating system hin-drances. We describe them in detail and present solutions for them. We use NAMD, a molecular dynamics program, as a case study for many of our performance optimizations.
Thesis (Ph. D.)--University of Washington, 1997Two recent trends are affecting the design of medium-...
Multicomputer (distributed memory MIMD machines) have emerged as inexpensive, yet powerful parallel...
Several communication systems that claim to support high performance computing in clusters focus on ...
The current trends in high performance computing show that large machines with tens of thousands of ...
Advances in multiprocessor interconnect technology are leading to high performance networks. However...
Improving the performance of future computing systems will be based upon the ability of increasing t...
Technology trends suggest that future machines will relyon parallelism to meet increasing performanc...
Technology trends suggest that future machines will rely on parallelism to meet increasing performan...
Conventional wisdom suggests that the most efficient use of modern computing clusters employs techni...
The goal of the QPACE project is to build a novel cost-efficient massive parallel supercomputer opti...
Inter-processor communication in interconnectionnetworks can create bottlenecks for scalable paralle...
Governments, universities, and companies expend vast resources building the top supercomputers. The...
The current trends in high performance computing show that large machines with tens of thousands of ...
The efficient implementation of collective communication patterns in a parallel machine is a challen...
Future manycore Systems-on-Chip will integrate tens or even hundreds of cores. Tiled architectures h...
Thesis (Ph. D.)--University of Washington, 1997Two recent trends are affecting the design of medium-...
Multicomputer (distributed memory MIMD machines) have emerged as inexpensive, yet powerful parallel...
Several communication systems that claim to support high performance computing in clusters focus on ...
The current trends in high performance computing show that large machines with tens of thousands of ...
Advances in multiprocessor interconnect technology are leading to high performance networks. However...
Improving the performance of future computing systems will be based upon the ability of increasing t...
Technology trends suggest that future machines will relyon parallelism to meet increasing performanc...
Technology trends suggest that future machines will rely on parallelism to meet increasing performan...
Conventional wisdom suggests that the most efficient use of modern computing clusters employs techni...
The goal of the QPACE project is to build a novel cost-efficient massive parallel supercomputer opti...
Inter-processor communication in interconnectionnetworks can create bottlenecks for scalable paralle...
Governments, universities, and companies expend vast resources building the top supercomputers. The...
The current trends in high performance computing show that large machines with tens of thousands of ...
The efficient implementation of collective communication patterns in a parallel machine is a challen...
Future manycore Systems-on-Chip will integrate tens or even hundreds of cores. Tiled architectures h...
Thesis (Ph. D.)--University of Washington, 1997Two recent trends are affecting the design of medium-...
Multicomputer (distributed memory MIMD machines) have emerged as inexpensive, yet powerful parallel...
Several communication systems that claim to support high performance computing in clusters focus on ...