Evaluation and measurement of multiprocessor systems with shared memory

This study is concerned with several proposals concerning multiprocessor systems and with the various possible methods of evaluating such proposals. After a discussion of the advantages and disadvantages of several performance evaluation tools, the author decides that simulation is the only tool powerful enough to develop a model which would be of practical use, in the design, comparison and extension of systems. The main aims of the simulation package developed as part of this study are cost effectiveness, ease of use and generality. The methodology on which the simulation package is based is described in detail. The fundamental principles are that model design should reflect actual systems design, that measuring procedures should be carried out alongside design that models should be well documented and easily adaptable and that models should be dynamic. The simulation package itself is modular, and in this way reflects current design trends. This approach also aids documentation and ensures that the model is easily adaptable. It contains a skeleton structure and a library of segments which can be added to or directly swapped with segments of the skeleton structure, to form a model which fits a user's requirements. The study also contains the results of some experimental work carried out using the model, the first part of which tests• the model's capabilities by simulating a large operating system, the ICL George 3 system; the second part deals with general questions and some of the many proposals concerning multiprocessor systems.

The influence of memory in a threshold model for distributed task assignment

A nature inspired decentralised multi-agent algorithm is proposed to solve a problem of distributed task selection in which cities produce and store batches of different mail types. Agents must collect and process the mail batches, without a priori knowledge of the available mail at the cities or inter-agent communication. In order to process a different mail type than the previous one, agents must undergo a change-over during which it remains inactive. We propose a threshold based algorithm in order to maximise the overall efficiency (the average amount of mail collected). We show that memory, i.e. the possibility for agents to develop preferences for certain cities, not only leads to emergent cooperation between agents, but also to a significant increase in efficiency (above the theoretical upper limit for any memoryless algorithm), and we systematically investigate the influence of the various model parameters. Finally, we demonstrate the flexibility of the algorithm to changes in circumstances, and its excellent scalability.

Knowledge transfer in globally distributed teams:the role of transactive memory

This paper explores the role of transactive memory in enabling knowledge transfer between globally distributed teams. While the information systems literature has recently acknowledged the role transactive memory plays in improving knowledge processes and performance in colocated teams, little is known about its contribution to distributed teams. To contribute to filling this gap, knowledge-transfer challenges and processes between onsite and offshore teams were studied at TATA Consultancy Services. In particular, the paper describes the transfer of knowledge between onsite and offshore teams through encoding, storing and retrieving processes. An in-depth case study of globally distributed software development projects was carried out, and a qualitative, interpretive approach was adopted. The analysis of the case suggests that in order to overcome differences derived from the local contexts of the onsite and offshore teams (e.g. different work routines, methodologies and skills), some specific mechanisms supporting the development of codified and personalized ‘directories’ were introduced. These include the standardization of templates and methodologies across the remote sites as well as frequent teleconferencing sessions and occasional short visits. These mechanisms contributed to the development of the notion of ‘who knows what’ across onsite and offshore teams despite the challenges associated with globally distributed teams, and supported the transfer of knowledge between onsite and offshore teams. The paper concludes by offering theoretical and practical implications.

The role of prior exposure on the capture of attention by items in working memory

The Biased Competition Model (BCM) suggests both top-down and bottom-up biases operate on selective attention (e.g., Desimone & Duncan, 1995). It has been suggested that top-down control signals may arise from working memory. In support, Downing (2000) found faster responses to probes presented in the location of stimuli held vs. not held in working memory. Soto, Heinke, Humphreys, and Blanco (2005) showed the involuntary nature of this effect and that shared features between stimuli were sufficient to attract attention. Here we show that stimuli held in working memory had an influence on the deployment of attentional resources even when: (1) It was detrimental to the task, (2) there was equal prior exposure, and (3) there was no bottom-up priming. These results provide further support for involuntary top-down guidance of attention from working memory and the basic tenets of the BCM, but further discredit the notion that bottom-up priming is necessary for the effect to occur.

Distributed routing algorithms for interconnected FDDI LANS

The Fibre Distributed Data Interface (FDDI) represents the new generation of local area networks (LANs). These high speed LANs are capable of supporting up to 500 users over a 100 km distance. User traffic is expected to be as diverse as file transfers, packet voice and video. As the proliferation of FDDI LANs continues, the need to interconnect these LANs arises. FDDI LAN interconnection can be achieved in a variety of different ways. Some of the most commonly used today are public data networks, dial up lines and private circuits. For applications that can potentially generate large quantities of traffic, such as an FDDI LAN, it is cost effective to use a private circuit leased from the public carrier. In order to send traffic from one LAN to another across the leased line, a routing algorithm is required. Much research has been done on the Bellman-Ford algorithm and many implementations of it exist in computer networks. However, due to its instability and problems with routing table loops it is an unsatisfactory algorithm for interconnected FDDI LANs. A new algorithm, termed ISIS which is being standardized by the ISO provides a far better solution. ISIS will be implemented in many manufacturers routing devices. In order to make the work as practical as possible, this algorithm will be used as the basis for all the new algorithms presented. The ISIS algorithm can be improved by exploiting information that is dropped by that algorithm during the calculation process. A new algorithm, called Down Stream Path Splits (DSPS), uses this information and requires only minor modification to some of the ISIS routing procedures. DSPS provides a higher network performance, with very little additional processing and storage requirements. A second algorithm, also based on the ISIS algorithm, generates a massive increase in network performance. This is achieved by selecting alternative paths through the network in times of heavy congestion. This algorithm may select the alternative path at either the originating node, or any node along the path. It requires more processing and memory storage than DSPS, but generates a higher network power. The final algorithm combines the DSPS algorithm with the alternative path algorithm. This is the most flexible and powerful of the algorithms developed. However, it is somewhat complex and requires a fairly large storage area at each node. The performance of the new routing algorithms is tested in a comprehensive model of interconnected LANs. This model incorporates the transport through physical layers and generates random topologies for routing algorithm performance comparisons. Using this model it is possible to determine which algorithm provides the best performance without introducing significant complexity and storage requirements.

Laterality differences in memory and attention

A critical review of the auditory selective attention literature is presented, particular reference is made to methodological issues arising from the asymmetrical hemispheric representation of language in the context of the dominant research technique dichotic shadowing. Subsequently the concept of cerebral localization is introduced, and the experimental literature with reference to models of laterality effects in speech and audition discussed. The review indicated the importance of hemispheric asymmetries insofar as they might influence the results of dichotic shadowing tasks. It is suggested that there is a potential overlap between models of selective attention and hemispheric differences. In Experiment I, ~ a key experiment in auditory selective attention is replicated and by exercising control over possible laterality effects some of the conflicting results of earlier studies were reconciled. The three subsequent experiments, II, III and IV, are concerned with the recall of verbally shadowed inputs. A highly significant and consistent effect of ear of arrival upon the serial position of items recalled is reported. Experiment V is directed towards an analysis of the effect that the processing of unattended inputs has upon the serial position of attended items that are recalled. A significant effect of the type of unattended material upon the recall of attended items was found to be influenced by the ear of arrival of inputs. In Experiment VI, differences between the two ears as attended and unattended input channels were clarified. Two main conclusions were drawn from this work. First, that the dichotic shadowing technique cannot control attention. Instead the task aprocessing both channels of dichotic inputs is unevenly shared bet\'reen the hemispheres as a function of the ear shadowed. Consequently, evidence for the processing of unattended information is considered in terms of constraints imposed by asymmetries in the functional organization of language, not in terms of a limited processing capacity model. The second conclusion to be drawn is that laterality differences can be effectively examined using the dichotic shadowing technique, a new model of laterality differences is proposed and discussed.

Dynamic causal modeling of load-dependent modulation of effective connectivity within the verbal working memory network

Neuroimaging studies have consistently shown that working memory (WM) tasks engage a distributed neural network that primarily includes the dorsolateral prefrontal cortex, the parietal cortex, and the anterior cingulate cortex. The current challenge is to provide a mechanistic account of the changes observed in regional activity. To achieve this, we characterized neuroplastic responses in effective connectivity between these regions at increasing WM loads using dynamic causal modeling of functional magnetic resonance imaging data obtained from healthy individuals during a verbal n-back task. Our data demonstrate that increasing memory load was associated with (a) right-hemisphere dominance, (b) increasing forward (i.e., posterior to anterior) effective connectivity within the WM network, and (c) reduction in individual variability in WM network architecture resulting in the right-hemisphere forward model reaching an exceedance probability of 99% in the most demanding condition. Our results provide direct empirical support that task difficulty, in our case WM load, is a significant moderator of short-term plasticity, complementing existing theories of task-related reduction in variability in neural networks. Hum Brain Mapp, 2013. © 2013 Wiley Periodicals, Inc.

An improved memory management scheme for large scale graph computing engine GraphChi

GraphChi is the first reported disk-based graph engine that can handle billion-scale graphs on a single PC efficiently. GraphChi is able to execute several advanced data mining, graph mining and machine learning algorithms on very large graphs. With the novel technique of parallel sliding windows (PSW) to load subgraph from disk to memory for vertices and edges updating, it can achieve data processing performance close to and even better than those of mainstream distributed graph engines. GraphChi mentioned that its memory is not effectively utilized with large dataset, which leads to suboptimal computation performances. In this paper we are motivated by the concepts of 'pin ' from TurboGraph and 'ghost' from GraphLab to propose a new memory utilization mode for GraphChi, which is called Part-in-memory mode, to improve the GraphChi algorithm performance. The main idea is to pin a fixed part of data inside the memory during the whole computing process. Part-in-memory mode is successfully implemented with only about 40 additional lines of code to the original GraphChi engine. Extensive experiments are performed with large real datasets (including Twitter graph with 1.4 billion edges). The preliminary results show that Part-in-memory mode memory management approach effectively reduces the GraphChi running time by up to 60% in PageRank algorithm. Interestingly it is found that a larger portion of data pinned in memory does not always lead to better performance in the case that the whole dataset cannot be fitted in memory. There exists an optimal portion of data which should be kept in the memory to achieve the best computational performance.

Benchmarking of distributed computing engines spark and GraphLab for big data analytics

In this paper we evaluate and compare two representativeand popular distributed processing engines for large scalebig data analytics, Spark and graph based engine GraphLab. Wedesign a benchmark suite including representative algorithmsand datasets to compare the performances of the computingengines, from performance aspects of running time, memory andCPU usage, network and I/O overhead. The benchmark suite istested on both local computer cluster and virtual machines oncloud. By varying the number of computers and memory weexamine the scalability of the computing engines with increasingcomputing resources (such as CPU and memory). We also runcross-evaluation of generic and graph based analytic algorithmsover graph processing and generic platforms to identify thepotential performance degradation if only one processing engineis available. It is observed that both computing engines showgood scalability with increase of computing resources. WhileGraphLab largely outperforms Spark for graph algorithms, ithas close running time performance as Spark for non-graphalgorithms. Additionally the running time with Spark for graphalgorithms over cloud virtual machines is observed to increaseby almost 100% compared to over local computer clusters.