Design and performance analysis of load-distributing fault-tolerant network

High-performance supercomputers generally comprise millions of CPUs in which interconnection networks play an important role to achieve high performance. New design paradigms of dynamic on-chip interconnection network involve a) topology b) synthesis, modeling and evaluation c) quality of service, fault tolerance and reliability d) routing procedures. To construct a dynamic highly fault tolerant interconnection networks requires more disjoint paths from each source-destination node pair at each stage and dynamic rerouting capability to use the various available paths effectively. Fast routing and rerouting strategy is needed to provide reliable performance on switch/link failures. This paper proposes two new architecture designs of fault tolerant interconnection networks named as reliable interconnection networks (RIN-1 and RIN-2). The proposed layouts are multipath multi-stage interconnection networks providing four disjoint paths for all the source-destination node pairs with dynamic rerouting capability. The designs can withstand switch failures in all the stages (including input and output stages) and provide more reliability. Reliability analysis of various MIN architectures is evaluated. On comparing the results with some existing MINs it is evident that the proposed designs provides higher reliability values and fault tolerance.

Performance analysis of synchronous multibuffered packet-switching networks

Multistage interconnection networks are used as a medium for interconnecting processors and memories in multiprocessor systems. Multistage interconnection networks are an effective substitute for crossbar switches, used in telephone network systems and parallel computers, due to their low cost, better performance and easy maintainability. In this paper, we propose new fault-tolerant hybrid multistage interconnection networks obtained from shuffle exchange gamma interconnection network (SEGIN) and logical neighborhood network named as shuffle exchange gamma logical neighborhood interconnection network (SEGLNIN). We evaluated the performance of SEGLNIN in terms of disjoint paths, reliability and hardware cost and compared it with some well-known MINs like shuffle exchange network (SEN), shuffle exchange network with one extra stage (SEN +) and shuffle exchange network with two extra stages (SEN + 2), SEGIN-1, and SEGIN-2. The results illustrate that the performance of the proposed SEGLNIN is better than that of the compared networks in terms of reliability and disjoint paths.

Interconnection networks: Physical design andperformance analysis

The conscious manipulation of mental representations is central to many creative and uniquely human abilities. How does the human brain mediate such flexible mental operations? Here, multivariate pattern analysis of functional MRI data reveals a widespread neural network that performs specific mental manipulations on the contents of visual imagery. Evolving patterns of neural activity within this mental workspace track the sequence of informational transformations carried out by these manipulations. The network switches between distinct connectivity profiles as representations are maintained or manipulated.

The design of packet radio networks involves a large number of issues which interact in a very complex fashion. Many of these pertain to the RF channel and its use, others pertain to the operational protocols. Clearly, no single model can be formulated which incorporates all the necessary parameters and leads to the optimum solution. The one essential element which complicates matters is that, contrary to point-to-point networks in which each channel is utilized by a single pair of nodes, the radio channel in packet radio networks is a multiaccess broadcast resource: i) in a given locality determined by radio connectivity, the channel is shared by many contending users, hence the need for channel access protocols; ii) radio is a broadcast medium and thus the action taken by a node has an effect on the actions taken by neighboring nodes and their outcome. Despite the complexity of the problem, there has been significant progress worth reporting on. The work accomplished so far has been either the analysis of specific examples of networks or an attempt to create models that would be useful in the design of general networks. The purpose of this paper is to survey the various modeling techniques that have been used for the performance analysis of packet radio networks, and to discuss the assumptions underlying these models, their scope of applicability, and some of the results obtained.

Cascading failures in interconnected networks with dynamical redistribution of loads

he analysis of any hardware or software from the point of view of the actual and possible performance is crucial in the computer world. Huge efforts are made to create newer, better and faster computers and programs while minor design flaws can result in very low performance not even worth the invested money. Many scientists focus their research on the performance analysis of different areas like parallel programs, PC clusters, shared memory architectures, supercomputers, cache performance and so on. Basically, two approaches can be found in this effort. One group of researchers try to create analitical models to calculate the performance and limits “in theory” while the other group implements monitoring tools to collect as much data about real executions as they can and then process this data and find bottlenecks in the performance. The first session of the workshop deals with performance analysis issues of different areas in computer science. The performance of PC clusters is analysed here as well as parallel applications with the need of the power of a teraflop supercomputer. Simulator for low-level interconnection networks is presented as well as mathematical models for parallel programs. Every researcher aims at making possible to the users to improve the performance of their system and their applications. The first paper is “Beowulf Performance in CFD Multigrid Applications” authored by F. Tirado, C. Garcia, R.S. Montereo, M. Prieto and I.M. Llorente. Authors evaluated the performance of a Beowulf-class system from the viewpoint of Computational Fluid Dynamics applications. They used two multigrid solvers to test their PC cluster with different configurations. The results show that, not surprisingly, the best configuration depends on the target application. The second paper presents “SICOSYS: An Integrated Framework for studying Interconnection Network Performance in Multiprocessor Systems” written by V. Puente, J.A. Gregorio and R. Beivide. SICOSYS is a generalpurpose simulator that makes possible the analysis of the interconnection networks of multiprocessor systems. This simulator has been integrated into the RSIM shared-memory-architecture and processor simulator as well as into the SimOS simulator that also includes the effects of an operating system into the simulation. The third paper presents new ideas about “Performance Analysis for Teraflop Computers: A Distributed Automatic Approach” authored by M. Gerndt, A. Schmidt, M. Schulz and R. Wismuller. A new performance analysis system has been designed for the monitoring and analysis of extremely computationally intensive applications on a teraflop computer. In such environment, classical analysis techniques fail because they are not scalable for thousands of processors. This paper presents the design of a new, automatic performance tool that is scalable on teraflop computers using a distributed approach for the analysis process. The fourth paper is the “Execution Time Prediction for Parallel Data Processing Tasks” written by S. Juhasz. This paper presents a mathematical model for execution time prediction in the domain of parallel programs with moderate computational power needs but working on large amount of data. A parallel sorting application is then analysed as an example.

By increasing the number of processors and with the help of the large-scale interconnection networks built on processors, the advanced multiprocessor systems, obtain their superpower of computing. However, the large-scale interconnection networks are very significant to the performance of any multiprocessor systems, but it also makes multiprocessor systems difficult to diagnose and isolate faulty processors. That’s why the multiprocessor systems should have a better fault diagnosis ability to ensure its reliability, and the ability of fault diagnosis of a multiprocessor system is normally measured by its diagnosability. Lexicographic product is a common method to construct a large scale interconnection network, so studying the diagnosability of multiprocessor system generated by the lexicographic product of networks will be beneficial to the development of multiprocessor systems. In this paper, we use the extending stars to obtain the accurate value of the diagnosability of the interconnection networks constructed by lexicographic product of m-order paths with n-order cycles, and prove that the diagnosability of lexicographic product of network is better than the Cartesian product of networks in certain cases.

Performance and dependability analysis of fault-tolerant networks

A new family of dynamic circuit switching Interconnection Networks is proposed, for connecting resources in parallel processing systems. These networks provide links for any input to any output, under distributed control schemes. Network nodes are build by 2×4/4×2, switches at the first/last stage, and 4×4 nodes in the middle stages. The networks compare favourably to existing non-blocking and blocking networks in terms of area and speed.

Performance analysis of interconnection networks for Wolf architecture is provided. Based on communication requirements, an optical interconnection network model for the architecture, an architectural modification, a new model for the Wolf architecture were suggested to investigate the possibilities for performance gain. Both architecture models were simulated with SAW simulator. SAW is a time-driven simulator that implements the Wolf characteristics. Parameters such as load distribution and execution time were dealt with so that performance of the architecture might be investigated. Results by simulations showed that the new model of Wolf architecture gives better results for the parameters investigated

In this chapter, the author will first introduce in Section 4.1 a brief overview of various approaches and processes for network analysis, architecture, and design. Details of the network analysis, architecture, and design processes will be sequentially presented in Sections 4.2-4.4.As a typical network-design model, the hierarchical design approach will be discussed in detail in Section 4.5 along with its layers in Section 4.6. Various network-design approaches and their evolution will be summarized in Section 4.7. Finally, Section 4.8 will provide further discussion on network management and security.

A fault tolerant multistage interconnection network with partly duplicated switches

A system designed for supporting the network performance analysis and forecast effort is presented, based on the combination of offline network analysis and online real-time performance forecast. The off-line analysis will perform analysis of specific network node performance, correlation analysis of relative network nodes performance and evolutionary mathematical modeling of long-term network performance measurements. The online real-time network performance forecast will be based on one so-called hybrid prediction modeling approach for short-term network, performance prediction and trend analysis. Based on the module design, the system proposed has good intelligence, scalability and self-adaptability, which will offer highly effective network performance analysis and forecast tools for network managers, and is one ideal support platform for network performance analysis and forecast effort.