MUHAMMAD K A L A M U D D I N AHAMAD (original) (raw)

Thesis Chapters by MUHAMMAD K A L A M U D D I N AHAMAD

Research paper thumbnail of A Framework for Embedded Hypercube Interconnection Networks: Based on Neural Network Approach

International Journal of Computer Applications, 2015

This paper is concerned with routing of data in an embedded hypercube interconnection using the a... more This paper is concerned with routing of data in an embedded hypercube interconnection using the approach based on neural net architecture. To present a framework of the interconnection network consist number of nodes and number of connections. In this paper we first show that n dimensional hypercube can be embedded in layer neural layer network such that for any node of hypercube, its neighboring nodes of other layer are evenly partition into layers where each layer shares a manipulating or resulting data of different layers. Under this embedding network to fixed target and varying data input to produce output of the two incidence matrix of kary n-cube network to embedded in architecture.

Papers by MUHAMMAD K A L A M U D D I N AHAMAD

Research paper thumbnail of Literature Review: Convey the Data in Massive Parallel Computing

International Journal of Innovative Research in Information Security (IJIRIS), 2015

In this paper we have studied several works on direct network architectures which are well-built ... more In this paper we have studied several works on direct network architectures which are well-built contestant for useful in many successful cost-effective, experimental massive parallel computers and well scale up shared memory of multiprocessors. The uniqueness of direct networks, as reflected by the communication latency and routing latency metrics are significant to the performance of such systems. A multiprocessor system can be used for the wormhole routing for the most capable switching method and has been adopted in several new massive parallel computers. This technique is unique technical challenges in routing and flow control in particular system, and avoid deadlock. The highly scale up network is a combination of topology and hypercube. Due to the being of concurrent multiple mesh and hypercubes, this network provides a great architectural support for parallel processing. The growth of the network is more efficient in terms of communication, interconnection network is scaled up the network and will be more reliable and also the unreliability of the interconnection network to get minimized. This is very desirable characteristic for the interconnection network as the network remains equipped for more failure of adjoining nodes or links in parallel computer architecture. Formulations to optimize the performance of throughput of networks through queuing theory M\M\1 concept.

Research paper thumbnail of A Statistical Analysis and Comparative Study of Embedded Hypercube

International Journal of Computer Applications, 2014

Consider the K-arry n-cube network is the most significant network structure in parallel computer... more Consider the K-arry n-cube network is the most significant network structure in parallel computer architecture. Therefore the generic structure of K-arry n-cube used to design the various networks with static network topology of different parameters, and its makes different types hypercube networks (N=2 n) from static networks known as embedded hypercube scalable inter-PE connection network suitable for parallel computing systems. Show how to design the good interconnection network in parallel Architecture for less density as well as diameter to route the data to different path. Also it has been proved with the computational results of entire system that the embedded hypercube interconnection networks built is highly scale up in terms of communication. A complete design analysis and comparison of network with various other networks is given using different network parameters, and statistical analysis comparative optimal of torus architecture rather than mesh architecture.

Research paper thumbnail of A Framework for Embedded Hypercube Interconnection Networks: Based on Neural Network Approach

International Journal of Computer Applications, 2015

This paper is concerned with routing of data in an embedded hypercube interconnection using the a... more This paper is concerned with routing of data in an embedded hypercube interconnection using the approach based on neural net architecture. To present a framework of the interconnection network consist number of nodes and number of connections. In this paper we first show that n dimensional hypercube can be embedded in layer neural layer network such that for any node of hypercube, its neighboring nodes of other layer are evenly partition into layers where each layer shares a manipulating or resulting data of different layers. Under this embedding network to fixed target and varying data input to produce output of the two incidence matrix of kary n-cube network to embedded in architecture.

Research paper thumbnail of Literature Review: Convey the Data in Massive Parallel Computing

International Journal of Innovative Research in Information Security (IJIRIS), 2015

In this paper we have studied several works on direct network architectures which are well-built ... more In this paper we have studied several works on direct network architectures which are well-built contestant for useful in many successful cost-effective, experimental massive parallel computers and well scale up shared memory of multiprocessors. The uniqueness of direct networks, as reflected by the communication latency and routing latency metrics are significant to the performance of such systems. A multiprocessor system can be used for the wormhole routing for the most capable switching method and has been adopted in several new massive parallel computers. This technique is unique technical challenges in routing and flow control in particular system, and avoid deadlock. The highly scale up network is a combination of topology and hypercube. Due to the being of concurrent multiple mesh and hypercubes, this network provides a great architectural support for parallel processing. The growth of the network is more efficient in terms of communication, interconnection network is scaled up the network and will be more reliable and also the unreliability of the interconnection network to get minimized. This is very desirable characteristic for the interconnection network as the network remains equipped for more failure of adjoining nodes or links in parallel computer architecture. Formulations to optimize the performance of throughput of networks through queuing theory M\M\1 concept.

Research paper thumbnail of A Statistical Analysis and Comparative Study of Embedded Hypercube

International Journal of Computer Applications, 2014

Consider the K-arry n-cube network is the most significant network structure in parallel computer... more Consider the K-arry n-cube network is the most significant network structure in parallel computer architecture. Therefore the generic structure of K-arry n-cube used to design the various networks with static network topology of different parameters, and its makes different types hypercube networks (N=2 n) from static networks known as embedded hypercube scalable inter-PE connection network suitable for parallel computing systems. Show how to design the good interconnection network in parallel Architecture for less density as well as diameter to route the data to different path. Also it has been proved with the computational results of entire system that the embedded hypercube interconnection networks built is highly scale up in terms of communication. A complete design analysis and comparison of network with various other networks is given using different network parameters, and statistical analysis comparative optimal of torus architecture rather than mesh architecture.