Mapping Parallel Iterative Algorithms onto Workstation Networks

For communication-intensive parallel applications, the maximum degree of concurrency achievable is limited by the communication throughput made available by the network. In previous work [HPS94], we showed experimentally that the performance of certain parallel applications running on a workstation network can be improved significantly if a congestion control protocol is used to enhance network performance. In this paper, we characterize and analyze the communication requirements of a large class of supercomputing applications that fall under the category of fixed-point problems, amenable to solution by parallel iterative methods. This results in a set of interface and architectural features sufficient for the efficient implementation of the applications over a large-scale distributed system. In particular, we propose a direct link between the application and network layer, supporting congestion control actions at both ends. This in turn enhances the system's responsiveness to network congestion, improving performance. Measurements are given showing the efficacy of our scheme to support large-scale parallel computations.

Mermera: Non-Coherent Distributed Shared Memory for Parallel Computing

The proliferation of inexpensive workstations and networks has prompted several researchers to use such distributed systems for parallel computing. Attempts have been made to offer a shared-memory programming model on such distributed memory computers. Most systems provide a shared-memory that is coherent in that all processes that use it agree on the order of all memory events. This dissertation explores the possibility of a significant improvement in the performance of some applications when they use non-coherent memory. First, a new formal model to describe existing non-coherent memories is developed. I use this model to prove that certain problems can be solved using asynchronous iterative algorithms on shared-memory in which the coherence constraints are substantially relaxed. In the course of the development of the model I discovered a new type of non-coherent behavior called Local Consistency. Second, a programming model, Mermera, is proposed. It provides programmers with a choice of hierarchically related non-coherent behaviors along with one coherent behavior. Thus, one can trade-off the ease of programming with coherent memory for improved performance with non-coherent memory. As an example, I present a program to solve a linear system of equations using an asynchronous iterative algorithm. This program uses all the behaviors offered by Mermera. Third, I describe the implementation of Mermera on a BBN Butterfly TC2000 and on a network of workstations. The performance of a version of the equation solving program that uses all the behaviors of Mermera is compared with that of a version that uses coherent behavior only. For a system of 1000 equations the former exhibits at least a 5-fold improvement in convergence time over the latter. The version using coherent behavior only does not benefit from employing more than one workstation to solve the problem while the program using non-coherent behavior continues to achieve improved performance as the number of workstations is increased from 1 to 6. This measurement corroborates our belief that non-coherent shared memory can be a performance boon for some applications.

Distributed Parallel Computing in Mermera: Mixing Noncoherent Shared Memories

Programmers of parallel processes that communicate through shared globally distributed data structures (DDS) face a difficult choice. Either they must explicitly program DDS management, by partitioning or replicating it over multiple distributed memory modules, or be content with a high latency coherent (sequentially consistent) memory abstraction that hides the DDS' distribution. We present Mermera, a new formalism and system that enable a smooth spectrum of noncoherent shared memory behaviors to coexist between the above two extremes. Our approach allows us to define known noncoherent memories in a new simple way, to identify new memory behaviors, and to characterize generic mixed-behavior computations. The latter are useful for programming using multiple behaviors that complement each others' advantages. On the practical side, we show that the large class of programs that use asynchronous iterative methods (AIM) can run correctly on slow memory, one of the weakest, and hence most efficient and fault-tolerant, noncoherence conditions. An example AIM program to solve linear equations, is developed to illustrate: (1) the need for concurrently mixing memory behaviors, and, (2) the performance gains attainable via noncoherence. Other program classes tolerate weak memory consistency by synchronizing in such a way as to yield executions indistinguishable from coherent ones. AIM computations on noncoherent memory yield noncoherent, yet correct, computations. We report performance data that exemplifies the potential benefits of noncoherence, in terms of raw memory performance, as well as application speed.

Iterative Linearized Density Matrix Propagation for Modeling Coherent Energy Transfer in Photosynthetic Light Harvesting

We present results of calculations [1] that employ a new mixed quantum classical iterative density matrix propagation approach (ILDM , or so called Is‐Landmap) [2] to explore the survival of coherence in different photo synthetic models. Our model studies confirm the long lived quantum coherence , while conventional theoretical tools (such as Redfield equation) fail to describe these phenomenon [3,4]. Our ILDM method is a numerical exactly propagation scheme and can be served as a bench mark calculation tools[2]. Result get from ILDM and from other recent methods have been compared and show agreement with each other[4,5]. Long lived coherence plateau has been attribute to the shift of harmonic potential due to the system bath interaction, and the harvesting efficiency is a balance between the coherence and dissipation[1]. We use this approach to investigate the excitation energy transfer dynamics in various light harvesting complex include Fenna‐Matthews‐Olsen light harvesting complex[1] and Cryptophyte Phycocyanin 645 [6]. [1] P.Huo and D.F.Coker ,J. Chem. Phys. 133, 184108 (2010) . [2] E.R. Dunkel, S. Bonella, and D.F. Coker, J. Chem. Phys. 129, 114106 (2008). [3] A. Ishizaki and G.R. Fleming, J. Chem. Phys. 130, 234111 (2009). [4] A. Ishizaki and G.R. Fleming, Proc. Natl. Acad. Sci. 106, 17255 (2009). [5] G. Tao and W.H. Miller, J. Phys. Chem. Lett. 1, 891 (2010). [6] P.Huo and D.F.Coker in preparation

Pricing Algorithms For a Two-sided Internet Advertisement Market

The Google AdSense Program is a successful internet advertisement program where Google places contextual adverts on third-party websites and shares the resulting revenue with each publisher. Advertisers have budgets and bid on ad slots while publishers set reserve prices for the ad slots on their websites. Following previous modelling efforts, we model the program as a two-sided market with advertisers on one side and publishers on the other. We show a reduction from the Generalised Assignment Problem (GAP) to the problem of computing the revenue maximising allocation and pricing of publisher slots under a first-price auction. GAP is APX-hard but a (1-1/e) approximation is known. We compute truthful and revenue-maximizing prices and allocation of ad slots to advertisers under a second-price auction. The auctioneer's revenue is within (1-1/e) second-price optimal.

How Good are Genetic Algorithms at Finding Large Cliques: An Experimental Study

This paper investigates the power of genetic algorithms at solving the MAX-CLIQUE problem. We measure the performance of a standard genetic algorithm on an elementary set of problem instances consisting of embedded cliques in random graphs. We indicate the need for improvement, and introduce a new genetic algorithm, the multi-phase annealed GA, which exhibits superior performance on the same problem set. As we scale up the problem size and test on \hard" benchmark instances, we notice a degraded performance in the algorithm caused by premature convergence to local minima. To alleviate this problem, a sequence of modi cations are implemented ranging from changes in input representation to systematic local search. The most recent version, called union GA, incorporates the features of union cross-over, greedy replacement, and diversity enhancement. It shows a marked speed-up in the number of iterations required to find a given solution, as well as some improvement in the clique size found. We discuss issues related to the SIMD implementation of the genetic algorithms on a Thinking Machines CM-5, which was necessitated by the intrinsically high time complexity (O(n3)) of the serial algorithm for computing one iteration. Our preliminary conclusions are: (1) a genetic algorithm needs to be heavily customized to work "well" for the clique problem; (2) a GA is computationally very expensive, and its use is only recommended if it is known to find larger cliques than other algorithms; (3) although our customization e ort is bringing forth continued improvements, there is no clear evidence, at this time, that a GA will have better success in circumventing local minima.

On the Performance of Polynomial-time CLIQUE Approximation Algorithms on Very Large Graphs

The performance of a randomized version of the subgraph-exclusion algorithm (called Ramsey) for CLIQUE by Boppana and Halldorsson is studied on very large graphs. We compare the performance of this algorithm with the performance of two common heuristic algorithms, the greedy heuristic and a version of simulated annealing. These algorithms are tested on graphs with up to 10,000 vertices on a workstation and graphs as large as 70,000 vertices on a Connection Machine. Our implementations establish the ability to run clique approximation algorithms on very large graphs. We test our implementations on a variety of different graphs. Our conclusions indicate that on randomly generated graphs minor changes to the distribution can cause dramatic changes in the performance of the heuristic algorithms. The Ramsey algorithm, while not as good as the others for the most common distributions, seems more robust and provides a more even overall performance. In general, and especially on deterministically generated graphs, a combination of simulated annealing with either the Ramsey algorithm or the greedy heuristic seems to perform best. This combined algorithm works particularly well on large Keller and Hamming graphs and has a competitive overall performance on the DIMACS benchmark graphs.

A Spectrum of TCP-Friendly Window-Based Congestion Control Algorithms

The increased diversity of Internet application requirements has spurred recent interests in transport protocols with flexible transmission controls. In window-based congestion control schemes, increase rules determine how to probe available bandwidth, whereas decrease rules determine how to back off when losses due to congestion are detected. The parameterization of these control rules is done so as to ensure that the resulting protocol is TCP-friendly in terms of the relationship between throughput and loss rate. In this paper, we define a new spectrum of window-based congestion control algorithms that are TCP-friendly as well as TCP-compatible under RED. Contrary to previous memory-less controls, our algorithms utilize history information in their control rules. Our proposed algorithms have two salient features: (1) They enable a wider region of TCP-friendliness, and thus more flexibility in trading off among smoothness, aggressiveness, and responsiveness; and (2) they ensure a faster convergence to fairness under a wide range of system conditions. We demonstrate analytically and through extensive ns simulations the steady-state and transient behaviors of several instances of this new spectrum of algorithms. In particular, SIMD is one instance in which the congestion window is increased super-linearly with time since the detection of the last loss. Compared to recently proposed TCP-friendly AIMD and binomial algorithms, we demonstrate the superiority of SIMD in: (1) adapting to sudden increases in available bandwidth, while maintaining competitive smoothness and responsiveness; and (2) rapidly converging to fairness and efficiency.

A Spectrum of TCP-friendly Window-Based Congestion Control Algorithms

The increasing diversity of Internet application requirements has spurred recent interest in transport protocols with flexible transmission controls. In window-based congestion control schemes, increase rules determine how to probe available bandwidth, whereas decrease rules determine how to back off when losses due to congestion are detected. The control rules are parameterized so as to ensure that the resulting protocol is TCP-friendly in terms of the relationship between throughput and loss rate. This paper presents a comprehensive study of a new spectrum of window-based congestion controls, which are TCP-friendly as well as TCP-compatible under RED. Our controls utilize history information in their control rules. By doing so, they improve the transient behavior, compared to recently proposed slowly-responsive congestion controls such as general AIMD and binomial controls. Our controls can achieve better tradeoffs among smoothness, aggressiveness, and responsiveness, and they can achieve faster convergence. We demonstrate analytically and through extensive ns simulations the steady-state and transient behavior of several instances of this new spectrum.

Popularity-Aware GreedyDual-Size Web Proxy Caching Algorithms

Web caching aims to reduce network traffic, server load, and user-perceived retrieval delays by replicating "popular" content on proxy caches that are strategically placed within the network. While key to effective cache utilization, popularity information (e.g. relative access frequencies of objects requested through a proxy) is seldom incorporated directly in cache replacement algorithms. Rather, other properties of the request stream (e.g. temporal locality and content size), which are easier to capture in an on-line fashion, are used to indirectly infer popularity information, and hence drive cache replacement policies. Recent studies suggest that the correlation between these secondary properties and popularity is weakening due in part to the prevalence of efficient client and proxy caches (which tend to mask these correlations). This trend points to the need for proxy cache replacement algorithms that directly capture and use popularity information. In this paper, we (1) present an on-line algorithm that effectively captures and maintains an accurate popularity profile of Web objects requested through a caching proxy, (2) propose a novel cache replacement policy that uses such information to generalize the well-known GreedyDual-Size algorithm, and (3) show the superiority of our proposed algorithm by comparing it to a host of recently-proposed and widely-used algorithms using extensive trace-driven simulations and a variety of performance metrics.

Construction of Neural Network Classification Expert Systems Using Switching Theory Algorithms

A new family of neural network architectures is presented. This family of architectures solves the problem of constructing and training minimal neural network classification expert systems by using switching theory. The primary insight that leads to the use of switching theory is that the problem of minimizing the number of rules and the number of IF statements (antecedents) per rule in a neural network expert system can be recast into the problem of minimizing the number of digital gates and the number of connections between digital gates in a Very Large Scale Integrated (VLSI) circuit. The rules that the neural network generates to perform a task are readily extractable from the network's weights and topology. Analysis and simulations on the Mushroom database illustrate the system's performance.