A generalized method for evaluation of pipeline processors under more realistic conditions than has been previously considered is investigated. First, several alternatives for a performance measure for pipeline archit...
详细信息
A generalized method for evaluation of pipeline processors under more realistic conditions than has been previously considered is investigated. First, several alternatives for a performance measure for pipeline architectures, and an index is presented that clearly reflects how well the pipeline has been organized from a purely architectural point of view. A Markov Chain model is then proposed for pipeline processors, and a method is suggested for determining the performance factor. this is followed by an illustrative example and some results of a preliminary analysis of the Texas Instruments Advanced Scientific Computer.
Empirical results are presented on the performance of parallel computations, with respect to various performance criteria, under different assumptions of the underlying computer architecture. the performance criteria ...
详细信息
Empirical results are presented on the performance of parallel computations, with respect to various performance criteria, under different assumptions of the underlying computer architecture. the performance criteria used are the parallel Index, the Speedup, the Utilization, the Efficiency, the Redundancy, the Compression and a definition of the Quality of the resultant computation. the underlying architectures assumed are parallel processor organizations of boththe SIMD and MIMD varieties, with limited and unlimited degrees of physical parallelism.
this conference proceedings contains 61 papers, of which 25 papers are indexed separately. the topics covered are: software and languages;architecture;interconnections;performance;distributed processing;algorithms and...
详细信息
this conference proceedings contains 61 papers, of which 25 papers are indexed separately. the topics covered are: software and languages;architecture;interconnections;performance;distributed processing;algorithms and applications;and data base architecture and software.
Hardwired resource allocators for TRAC-like reconfigurable architectures are described. these allocators facilitate searching for available resources in the system and allocation of a subset of these to a given reques...
详细信息
Hardwired resource allocators for TRAC-like reconfigurable architectures are described. these allocators facilitate searching for available resources in the system and allocation of a subset of these to a given request. Various algorithms can be implemented for the search and the allocation of the resources. Tree-structured allocators look particularly attractive withthe cost-delay product being of the order of M* (log M)**2 for a system with M resources of the same type. It is shown how this scheme can be extended to allocate multiple type of resources in the system.
Empirical results are presented on the performance of parallel computations, with respect to various performance criteria, under different assumptions of the underlying computer architecture. the performance criteria ...
详细信息
Empirical results are presented on the performance of parallel computations, with respect to various performance criteria, under different assumptions of the underlying computer architecture. the performance criteria used are the parallel Index, the Speedup, the Utilization, the Efficiency, the Redundancy, the Compression and a definition of the Quality of the resultant computation. the underlying architectures assumed are parallel processor organizations of boththe SIMD and MIMD varieties, with limited and unlimited degrees of physical parallelism.
A generalized method for evaluation of pipeline processors under more realistic conditions than has been previously considered is investigated. First, several alternatives for a performance measure for pipeline archit...
详细信息
A generalized method for evaluation of pipeline processors under more realistic conditions than has been previously considered is investigated. First, several alternatives for a performance measure for pipeline architectures, and an index is presented that clearly reflects how well the pipeline has been organized from a purely architectural point of view. A Markov Chain model is then proposed for pipeline processors, and a method is suggested for determining the performance factor. this is followed by an illustrative example and some results of a preliminary analysis of the Texas Instruments Advanced Scientific Computer.
the use of the SIMD mode of parallelism to perform the speech analysis task of linear predictive coding is explored. Linear prediction represents one of the major analysis techniques for speech compression, transmissi...
详细信息
the use of the SIMD mode of parallelism to perform the speech analysis task of linear predictive coding is explored. Linear prediction represents one of the major analysis techniques for speech compression, transmission, and recognition applications. parallelalgorithms to perform linear prediction have been developed, and are evaluated in terms of the number of arithmetic operations and interprocessor data transfers needed. From the algorithms, architectural requirements such as machine size and interconnection network capability are analyzed.
this conference proceedings contains 61 papers, of which 25 papers are indexed separately. the topics covered are: software and languages; architecture; interconnections; performance; distributed processing; algorithm...
详细信息
this conference proceedings contains 61 papers, of which 25 papers are indexed separately. the topics covered are: software and languages; architecture; interconnections; performance; distributed processing; algorithms and applications; and data base architecture and software.
the use of the SIMD mode of parallelism to perform the speech analysis task of linear predictive coding is explored. Linear prediction represents one of the major analysis techniques for speech compression, transmissi...
详细信息
the use of the SIMD mode of parallelism to perform the speech analysis task of linear predictive coding is explored. Linear prediction represents one of the major analysis techniques for speech compression, transmission, and recognition applications. parallelalgorithms to perform linear prediction have been developed, and are evaluated in terms of the number of arithmetic operations and interprocessor data transfers needed. From the algorithms, architectural requirements such as machine size and interconnection network capability are analyzed.
this book constitutes the refereed proceedings of the 19thinternationalconference on Advanced Data Mining and Applications, ADMA 2023, held in Shenyang, China, during August 21–23, 2023.;the 216 ...
详细信息
ISBN:
(数字)9783031466649
ISBN:
(纸本)9783031466632
this book constitutes the refereed proceedings of the 19thinternationalconference on Advanced Data Mining and Applications, ADMA 2023, held in Shenyang, China, during August 21–23, 2023.;the 216 full papers included in this book were carefully reviewed and selected from 503 submissions. they were organized in topical sections as follows: Data mining foundations, Grand challenges of data mining, parallel and distributed data mining algorithms, Mining on data streams, Graph mining and Spatial data mining.
暂无评论