Motivated by observations about job runtimes on the CPlant system, we use a trace-driven microsimulator to begin characterizing the performance of different classes of allocation algorithms on jobs with different communication patterns in space-shared parallel systems with mesh topology. We show that relative performance varies considerably with communication pattern. The Paging strategy using the Hilbert space-filling curve and the Best Fit heuristic performed best across several communication patterns
The performance of the existing non-contiguous processor allocation strategies has been traditionall...
The performance of the message-passing applications on a parallel system can vary and cause ine cien...
Researchers conducted experiments on parallel algorithms, which are inspired by the clonal selection...
Motivated by observations about job runtimes on the CPlant system, we use a trace-driven microsimula...
The Computational Plant or Cplant is a commodity-based distributed-memory supercomputer under develo...
International audienceMATCHING COMMUNICATION PATTERN WITH UNDERLYING HARDWARE ARCHITECTUR
This paper focuses on the performance of basic communication primitives, namely the overlap of messa...
Governments, universities, and companies expend vast resources building the top supercomputers. The...
In systems consisting of multiple clusters of processors which are interconnected by relatively slow...
Interprocessor communication overhead is a crucial measure of the power of parallel computing system...
We dedicate this article to the memory of Steve Seiden, who was killed in a tragic cycling accident ...
International audienceIn distributed memory systems, it is paramount to develop strategies to overla...
International audienceIt is well-known that taking into account communications while scheduling jobs...
scheduling In this paper, we utilize a bandwidth-centric job communication model that captures the i...
The problem of allocating nodes of a program graph to processors in a parallel processing architectu...
The performance of the existing non-contiguous processor allocation strategies has been traditionall...
The performance of the message-passing applications on a parallel system can vary and cause ine cien...
Researchers conducted experiments on parallel algorithms, which are inspired by the clonal selection...
Motivated by observations about job runtimes on the CPlant system, we use a trace-driven microsimula...
The Computational Plant or Cplant is a commodity-based distributed-memory supercomputer under develo...
International audienceMATCHING COMMUNICATION PATTERN WITH UNDERLYING HARDWARE ARCHITECTUR
This paper focuses on the performance of basic communication primitives, namely the overlap of messa...
Governments, universities, and companies expend vast resources building the top supercomputers. The...
In systems consisting of multiple clusters of processors which are interconnected by relatively slow...
Interprocessor communication overhead is a crucial measure of the power of parallel computing system...
We dedicate this article to the memory of Steve Seiden, who was killed in a tragic cycling accident ...
International audienceIn distributed memory systems, it is paramount to develop strategies to overla...
International audienceIt is well-known that taking into account communications while scheduling jobs...
scheduling In this paper, we utilize a bandwidth-centric job communication model that captures the i...
The problem of allocating nodes of a program graph to processors in a parallel processing architectu...
The performance of the existing non-contiguous processor allocation strategies has been traditionall...
The performance of the message-passing applications on a parallel system can vary and cause ine cien...
Researchers conducted experiments on parallel algorithms, which are inspired by the clonal selection...