We consider the problem of merging m disjoint ordered lists, each of size n/m. We determine up to a constant factor the worst case and average case deterministic and randomized parallel comparison complexity of the pr...
详细信息
We consider the problem of merging m disjoint ordered lists, each of size n/m. We determine up to a constant factor the worst case and average case deterministic and randomized parallel comparison complexity of the problem for all the range of n, m and p where p is the number of processors used. The worst case deterministic time complexity is [GRAPHICS] That means [GRAPHICS] and [GRAPHICS] Clearly merging two equal lists and sorting are special cases of this problem for m = 2 and m = n respectively. We also prove that these bounds hold for randomized algorithms and even for the average case of deterministic or randomized ones. Therefore the average case of the best deterministic or randomized algorithm for this problem is not faster than the worst case of the best deterministic one by more than a constant factor.
Efficient scheduling techniques of computing resources are essential for achieving satisfactory performance for users as computersystems and their applications become more complex. In this paper, we survey research o...
详细信息
Efficient scheduling techniques of computing resources are essential for achieving satisfactory performance for users as computersystems and their applications become more complex. In this paper, we survey research on scheduling algorithms, review previous classifications of scheduling problems, and present a broader classification scheme. Using a uniform terminology for scheduling strategies and the new classification scheme, previous work on scheduling strategies is reviewed and trends in scheduling research are identified. Finally, a methodology for developing scheduling strategies is presented.
We present a simple algorithm for emulating an N-processor CROW PRAM on an N-ode butterfly. Each step of the PRAM is emulated in time O(log N) with high probability, using FIFO queues of size O(1) at each node. The on...
详细信息
We present a simple algorithm for emulating an N-processor CROW PRAM on an N-ode butterfly. Each step of the PRAM is emulated in time O(log N) with high probability, using FIFO queues of size O(1) at each node. The only use of randomization is in selecting a hash function to distribute the shared address space of the PRAM onto the nodes of the butterfly. The routing itself is both deterministic and oblivious, and messages are combined without the use of associative memories or explicit sorting. As a corollary we improve the result of Pippenger by routing permutations with bounded queues in logarithmic time, without the possibility of deadlock. Besides being optimal, our algorithm has the advantage of extreme simplicity and is readily suited for use in practice.
This paper presents algorithms and their implementations for table look-up modeling of static and dynamic behavior of electronic devices for transient simulation. More specifically, multivariate Bernstein polynomials ...
详细信息
This paper presents algorithms and their implementations for table look-up modeling of static and dynamic behavior of electronic devices for transient simulation. More specifically, multivariate Bernstein polynomials are used to interpolate the operating point from tabular input-output data. For most device characteristics quadratic tensor product of input-output polynomial functions in Bernstein form offer operating point values within a few percent of the analytical function value. This range of accuracy is acceptable for most transient simulation scenarios. The algorithm outlined here consists of dot product evaluations and thus it is computationally simpler than analytical models.
A parallel version of quicksort on a CRCW PRAM is developed. The algorithm uses n processors and a linear space to sort n keys in the expected time O (log n ) with large probability.
A parallel version of quicksort on a CRCW PRAM is developed. The algorithm uses n processors and a linear space to sort n keys in the expected time O (log n ) with large probability.
This paper introduces a new concept regarding information processing in general, but more specifically inter-process communication (IPC) and semantic information processing, where a controlled environment is essential...
详细信息
This paper introduces a new concept regarding information processing in general, but more specifically inter-process communication (IPC) and semantic information processing, where a controlled environment is essential in order to overcome various problems such as deadlocks. The emphasis here is on communication management and the tools necessary for the user to control the IPC environment at both intra- and inter-process levels. The IPC system developed in C is also discussed in detail.
This paper presents methodologies capable of quantifying multiprogramming (MP) overhead on a computer system. Two methods which quantify the lower bound on MP overhead, along with a method to determine MP overhead pre...
详细信息
This paper presents methodologies capable of quantifying multiprogramming (MP) overhead on a computer system. Two methods which quantify the lower bound on MP overhead, along with a method to determine MP overhead present in real workloads, are introduced. The techniques are illustrated by determining the percentage of parallel processing time consumed by MP overhead on Alliant multiprocessors. The real workload MP overhead measurements, as well as measurements of other overhead components such as kernel lock spinning, are then used in a comprehensive case study of performance degradation due to overheads. It is found that MP overhead accounts for well over half of the total system overhead. Kernel lock spinning is determined to be a major component of both MP and total system overhead. Correlation analysis is used to uncover underlying relationships between overheads and workload characteristics. It is found that for the workloads studied, MP overhead in the parallel environment is not statistically dependent on the number of parallel jobs being multiprogrammed. However, because of increased kernel contention, serial jobs, even those executing on peripheral processors, are responsible for variation in MP overhead.
Two types of documentation delivery systems were examined. On-line and printed software documentation systems for a commercial data base management system were explored. In a controlled lab experiment, two groups of n...
详细信息
Two types of documentation delivery systems were examined. On-line and printed software documentation systems for a commercial data base management system were explored. In a controlled lab experiment, two groups of novice subjects using one or the other type of documentation were tested and the results compared. Printed-documentations users had better test scores and completed the test faster. On-line-documentation users were frustrated by inadequate instructions and examples. Overall, printed documentation was favored by the subjects in the study.
A newin situ external sorting algorithm, to be calledExtquick, is developed and its time and space performance are analysed. It is shown that Extquick performs more efficiently than similarin situ sorting algorithms b...
详细信息
A newin situ external sorting algorithm, to be calledExtquick, is developed and its time and space performance are analysed. It is shown that Extquick performs more efficiently than similarin situ sorting algorithms based on Quicksort that appear in the literature. Since the computational tree of Quicksort-like sorting algorithms is equivalent to a search tree, techniques that model the time complexity of such a structure are then used for the analysis of Extquick.
Task partitioning is an important technique in parallel *** this paper,we investigate the optimal partitioning strategies and granularities of tasks with communications based on several models of parallel computer ***...
详细信息
Task partitioning is an important technique in parallel *** this paper,we investigate the optimal partitioning strategies and granularities of tasks with communications based on several models of parallel computer *** from the usual approach,we study the optimal partitioning strate- gies and granularities from the viewpoint of minimizing T as well as minimizing NT^2,where N is the number of processors used and T is the program execution time using N *** results show that the optimal partitioning strategies for all cases discussed in this paper are the same——either to as- sign all tasks to one processor or to distribute them among the processors as equally as possible de- pending only on the functions of ratio of running time to communication time R/C.
暂无评论