This paper introduces the circulating processor model for parallel computer systems. The circulating processor model is a product form queuing network model where the processors are allowed to circulate between the parallel applications instead of the more traditional circulating task model. Certain behaviors of parallel systems are better captured using this new approach. The circulating processor model may be load dependent or load dependent. The load dependent circulating processor model is exact for systems which contain a single parallel application. An exact error is calculated for the load independent circulating processor model for systems which contain a single parallel application. The load dependent circulating processor model is...
The development of multiprocessor simulations from a serial set of ordinary differential equations d...
Contains fulltext : 240925.pdf (Publisher’s version ) (Open Access)16 p
The Parallel Processor Engine Model Program is a generalized engineering tool intended to aid in the...
A parallel programming archetype [Cha94, CMMM95] is an abstraction that captures the common features...
Many scientific applications for high-performance multiprocessors have tremendous I/O requirements. ...
Parallel disk systems provide opportunities for exploiting I/O parallelism in two possible ways, nam...
We study the parallel computation of dynamic programming. We consider four important dynamic program...
Finding the connected components of an undirected graph G=(V, E) on n=|V| vertices and m=|E| edges i...
It is easy to find errors and inefficient parts of a sequential program, by using a standard debugge...
The fast development over the last years of high performance multicomputers makes them attractive ca...
Here, we deal with a new fine grain parallel-pipelined architecture made up of heterogeneous N arith...
This paper presents a new parallel processing scheme called DYNAMIC-JOIN for OPS5-like production sy...
Although there are various multiprocessors available, both software support and algorithm developmen...
We show that defining a finite but unbounded number of parallel processes using the equation $S(k,dt...
The message passing programs are executed with the Parallel Virtual Machine (PVM) library and the sh...
The development of multiprocessor simulations from a serial set of ordinary differential equations d...
Contains fulltext : 240925.pdf (Publisher’s version ) (Open Access)16 p
The Parallel Processor Engine Model Program is a generalized engineering tool intended to aid in the...
A parallel programming archetype [Cha94, CMMM95] is an abstraction that captures the common features...
Many scientific applications for high-performance multiprocessors have tremendous I/O requirements. ...
Parallel disk systems provide opportunities for exploiting I/O parallelism in two possible ways, nam...
We study the parallel computation of dynamic programming. We consider four important dynamic program...
Finding the connected components of an undirected graph G=(V, E) on n=|V| vertices and m=|E| edges i...
It is easy to find errors and inefficient parts of a sequential program, by using a standard debugge...
The fast development over the last years of high performance multicomputers makes them attractive ca...
Here, we deal with a new fine grain parallel-pipelined architecture made up of heterogeneous N arith...
This paper presents a new parallel processing scheme called DYNAMIC-JOIN for OPS5-like production sy...
Although there are various multiprocessors available, both software support and algorithm developmen...
We show that defining a finite but unbounded number of parallel processes using the equation $S(k,dt...
The message passing programs are executed with the Parallel Virtual Machine (PVM) library and the sh...
The development of multiprocessor simulations from a serial set of ordinary differential equations d...
Contains fulltext : 240925.pdf (Publisher’s version ) (Open Access)16 p
The Parallel Processor Engine Model Program is a generalized engineering tool intended to aid in the...