Search results for: Long-short-term memory

Authors: Y. Bentoutou

Abstract:

On-board Error Detection and Correction (EDAC) devices aim to secure data transmitted between the central processing unit (CPU) of a satellite onboard computer and its local memory. This paper presents a comparison of the performance of four low complexity EDAC techniques for application in Random Access Memories (RAMs) on-board small satellites. The performance of a newly proposed EDAC architecture is measured and compared with three different EDAC strategies, using the same FPGA technology. A statistical analysis of single-event upset (SEU) and multiple-bit upset (MBU) activity in commercial memories onboard Alsat-1 is given for a period of 8 years

Keywords: Error Detection and Correction; On-board computer; small satellite missions

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2262

Authors: Heesung Lee, Joonkyung Sung, Euntai Kim

Abstract:

This paper proposes a method which reduces power consumption in single-error correcting, double error-detecting checker circuits that perform memory error correction code. Power is minimized with little or no impact on area and delay, using the degrees of freedom in selecting the parity check matrix of the error correcting codes. The genetic algorithm is employed to solve the non linear power optimization problem. The method is applied to two commonly used SEC-DED codes: standard Hamming and odd column weight Hsiao codes. Experiments were performed to show the performance of the proposed method.

Keywords: Error correcting codes, genetic algorithm, non-linearpower optimization, Hamming code, Hsiao code.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2185

Authors: Aytunc Durlanik, Ibrahim Sogukpinar

Abstract:

SIP (Session Initiation Protocol), using HTML based call control messaging which is quite simple and efficient, is being replaced for VoIP networks recently. As for authentication and authorization purposes there are many approaches and considerations for securing SIP to eliminate forgery on the integrity of SIP messages. On the other hand Elliptic Curve Cryptography has significant advantages like smaller key sizes, faster computations on behalf of other Public Key Cryptography (PKC) systems that obtain data transmission more secure and efficient. In this work a new approach is proposed for secure SIP authentication by using a public key exchange mechanism using ECC. Total execution times and memory requirements of proposed scheme have been improved in comparison with non-elliptic approaches by adopting elliptic-based key exchange mechanism.

Keywords: SIP, Elliptic Curve Cryptography, voice over IP.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2528

Authors: Wen-Bao Qiao, Jean-Charles Créput

Abstract:

To accelerate the solution for large scale traveling salesman problems (TSP), a parallel 2-opt local search algorithm with simple implementation based on Graphics Processing Unit (GPU) is presented and tested in this paper. The parallel scheme is based on technique of data decomposition by dynamically assigning multiple K processors on the integral tour to treat K edges’ 2-opt local optimization simultaneously on independent sub-tours, where K can be user-defined or have a function relationship with input size N. We implement this algorithm with doubly linked list on GPU. The implementation only requires O(N) memory. We compare this parallel 2-opt local optimization against sequential exhaustive 2-opt search along integral tour on TSP instances from TSPLIB with more than 10000 cities.

Keywords: Doubly linked list, parallel 2-opt, tour division, GPU.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1223

Authors: Earleen Jane Fuentes, Regeene Melarese Lim, Franklin Benjamin Tapia, Alexis Pantola

Abstract:

Mobile Ad-hoc Network (MANET) is an infrastructure-less network of mobile devices, also known as nodes. These nodes heavily rely on each other’s resources such as memory, computing power, and energy. Thus, some nodes may become selective in forwarding packets so as to conserve their resources. These nodes are called misleading nodes. Several reputation-based techniques (e.g. CORE, CONFIDANT, LARS, SORI, OCEAN) and acknowledgment-based techniques (e.g. TWOACK, S-TWOACK, EAACK) have been proposed to detect such nodes. These techniques do not appropriately punish misleading nodes. Hence, this paper addresses the limitations of these techniques using a system called MINDRA.

Keywords: Mobile ad-hoc network, selfish nodes, reputation-based techniques, acknowledgment-based techniques.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1376

Authors: Zhang Xiuping, Yang Guowu, Zheng Desheng

Abstract:

Embedded hardware simulator is a valuable computeraided tool for embedded application development. This paper focuses on the ARM926EJ-S MMU, builds state transition models and formally verifies critical properties for the models. The state transition models include loading instruction model, reading data model, and writing data model. The properties of the models are described by CTL specification language, and they are verified in VIS. The results obtained in VIS demonstrate that the critical properties of MMU are satisfied in the state transition models. The correct models can be used to implement the MMU component in our simulator. In the end of this paper, the experimental results show that the MMU can successfully accomplish memory access requests from CPU.

Keywords: MMU, State transition, Model, Simulation.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1617

Authors: Mario Mastriani

Abstract:

In this work, we developed the concept of supercompression, i.e., compression above the compression standard used. In this context, both compression rates are multiplied. In fact, supercompression is based on super-resolution. That is to say, supercompression is a data compression technique that superpose spatial image compression on top of bit-per-pixel compression to achieve very high compression ratios. If the compression ratio is very high, then we use a convolutive mask inside decoder that restores the edges, eliminating the blur. Finally, both, the encoder and the complete decoder are implemented on General-Purpose computation on Graphics Processing Units (GPGPU) cards. Specifically, the mentio-ned mask is coded inside texture memory of a GPGPU.

Keywords: General-Purpose computation on Graphics Processing Units, Image Compression, Interpolation, Super-resolution.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1979

Authors: Mohammad H. Fattahi

Abstract:

Chaotic analysis has been performed on the river flow time series before and after applying the wavelet based de-noising techniques in order to investigate the noise content effects on chaotic nature of flow series. In this study, 38 years of monthly runoff data of three gauging stations were used. Gauging stations were located in Ghar-e-Aghaj river basin, Fars province, Iran. Noise level of time series was estimated with the aid of Gaussian kernel algorithm. This step was found to be crucial in preventing removal of the vital data such as memory, correlation and trend from the time series in addition to the noise during de-noising process.

Keywords: Chaotic behavior, wavelet, noise reduction, river flow.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2095

Authors: Muazzam A. Khan, Ghalib A. Shah, Muhammad Sher

Abstract:

Wireless sensor network is formed with the combination of sensor nodes and sink nodes. Recently Wireless sensor network has attracted attention of the research community. The main application of wireless sensor network is security from different attacks both for mass public and military. However securing these networks, by itself is a critical issue due to many constraints like limited energy, computational power and lower memory. Researchers working in this area have proposed a number of security techniques for this purpose. Still, more work needs to be done.In this paper we provide a detailed discussion on security in wireless sensor networks. This paper will help to identify different obstacles and requirements for security of wireless sensor networks as well as highlight weaknesses of existing techniques.

Keywords: Wireless senor networks (WSNs), security, denial of service, black hole, cryptography, stenography.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2923

Authors: Zhenhuan Zhu, David Mulvaney, Vassilios Chouliaras

Abstract:

This paper presents a novel genetic algorithm, termed the Optimum Individual Monogenetic Algorithm (OIMGA) and describes its hardware implementation. As the monogenetic strategy retains only the optimum individual, the memory requirement is dramatically reduced and no crossover circuitry is needed, thereby ensuring the requisite silicon area is kept to a minimum. Consequently, depending on application requirements, OIMGA allows the investigation of solutions that warrant either larger GA populations or individuals of greater length. The results given in this paper demonstrate that both the performance of OIMGA and its convergence time are superior to those of existing hardware GA implementations. Local convergence is achieved in OIMGA by retaining elite individuals, while population diversity is ensured by continually searching for the best individuals in fresh regions of the search space.

Keywords: Genetic algorithms, hardware-based machinelearning.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1640

Authors: Daniel I. Morariu, Lucian N. Vintan, Volker Tresp

Abstract:

Text categorization is the problem of classifying text documents into a set of predefined classes. After a preprocessing step, the documents are typically represented as large sparse vectors. When training classifiers on large collections of documents, both the time and memory restrictions can be quite prohibitive. This justifies the application of feature selection methods to reduce the dimensionality of the document-representation vector. In this paper, we present three feature selection methods: Information Gain, Support Vector Machine feature selection called (SVM_FS) and Genetic Algorithm with SVM (called GA_SVM). We show that the best results were obtained with GA_SVM method for a relatively small dimension of the feature vector.

Keywords: Feature Selection, Learning with Kernels, Support Vector Machine, Genetic Algorithm, and Classification.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1706

Authors: Constantin Pehoiu, Cristian Savu, Silviu Badea, Cristian Borida

Abstract:

The physiological effects of physical exercise on human body are relatively well known in literature, which describes in detail the changes that occur in the cardiovascular system, the respiratory one, in bones and other systems, both during exercise and after its delivery. However, the effects of exercise on mental processes are less treated. From the literature reviews discussed in this study, it can be detached the idea that we can not exactly say that physical exercise has beneficial effects on mental processes, but neither that it would have potentially negative effects. This uncertainty, reflected in the inability to indicate precise and unequivocal meaning, favorable-unfavorable physical effort in acting on mental processes, is a prime reason to undertake a study of the phenomenon influence effort administered physical education classes on the dynamics of mental processes like attention and memory.

Keywords: management, exercise, mental process, lesson.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1666

Authors: Daniel Morariu, Lucian N. Vintan, Volker Tresp

Abstract:

Text categorization is the problem of classifying text documents into a set of predefined classes. After a preprocessing step, the documents are typically represented as large sparse vectors. When training classifiers on large collections of documents, both the time and memory restrictions can be quite prohibitive. This justifies the application of feature selection methods to reduce the dimensionality of the document-representation vector. In this paper, three feature selection methods are evaluated: Random Selection, Information Gain (IG) and Support Vector Machine feature selection (called SVM_FS). We show that the best results were obtained with SVM_FS method for a relatively small dimension of the feature vector. Also we present a novel method to better correlate SVM kernel-s parameters (Polynomial or Gaussian kernel).

Keywords: Feature Selection, Learning with Kernels, SupportVector Machine, and Classification.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1829

Authors: Po-Hsieh Lin, Jyi-Tsong Lin

Abstract:

A vertical SOI-based MOSFET with trench body structure operated as 1T DRAM cell at various temperatures has been studied and investigated. Different operation temperatures are assigned for the device for its performance comparison, thus the thermal stability is carefully evaluated for the future memory device applications. Based on the simulation, the vertical SOI-based MOSFET with trench body structure demonstrates the electrical characteristics properly and possess conspicuous kink effect at various operation temperatures. Transient characteristics were also performed to prove that its programming window values and retention time behaviors are acceptable when the new 1T DRAM cell is operated at high operation temperature.

Keywords: SOI, 1T DRAM, thermal stability.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1574

Authors: Marco Ryll, Svetan Ratchev

Abstract:

This paper discusses the applicability of the Data Distribution Service (DDS) for the development of automated and modular manufacturing systems which require a flexible and robust communication infrastructure. DDS is an emergent standard for datacentric publish/subscribe middleware systems that provides an infrastructure for platform-independent many-to-many communication. It particularly addresses the needs of real-time systems that require deterministic data transfer, have low memory footprints and high robustness requirements. After an overview of the standard, several aspects of DDS are related to current challenges for the development of modern manufacturing systems with distributed architectures. Finally, an example application is presented based on a modular active fixturing system to illustrate the described aspects.

Keywords: Flexible Manufacturing, Publish/Subscribe, Plug & Produce.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2352

Authors: Kyoung Kun Lee, Soongyu Kwon, Jong Tae Kim

Abstract:

An embedded system for SEU(single event upset) test needs to be designed to prevent system failure by high-energy particles during measuring SEU. SEU is a phenomenon in which the data is changed temporary in semiconductor device caused by high-energy particles. In this paper, we present an embedded system for SRAM(static random access memory) SEU test. SRAMs are on the DUT(device under test) and it is separated from control board which manages the DUT and measures the occurrence of SEU. It needs to have considerations for preventing system failure while managing the DUT and making an accurate measurement of SEUs. We measure the occurrence of SEUs from five different SRAMs at three different cyclotron beam energies 30, 35, and 40MeV. The number of SEUs of SRAMs ranges from 3.75 to 261.00 in average.

Keywords: embedded system, single event upset, SRAM

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1669

Authors: Wang Tong, He Pi-lian

Abstract:

This paper sets forth the possibility and importance about applying Data Mining in Web logs mining and shows some problems in the conventional searching engines. Then it offers an improved algorithm based on the original AprioriAll algorithm which has been used in Web logs mining widely. The new algorithm adds the property of the User ID during the every step of producing the candidate set and every step of scanning the database by which to decide whether an item in the candidate set should be put into the large set which will be used to produce next candidate set. At the meantime, in order to reduce the number of the database scanning, the new algorithm, by using the property of the Apriori algorithm, limits the size of the candidate set in time whenever it is produced. Test results show the improved algorithm has a more lower complexity of time and space, better restrain noise and fit the capacity of memory.

Keywords: Candidate Sets Pruning, Data Mining, ImprovedAlgorithm, Noise Restrain, Web Log

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2280

Authors: Bi Kun, Gu Nai-jie, Tu Kun, Liu Xiao-hu, Liu Gang

Abstract:

Traditional parallel single string matching algorithms are always based on PRAM computation model. Those algorithms concentrate on the cost optimal design and the theoretical speed. Based on the distributed string matching algorithm proposed by CHEN, a practical distributed string matching algorithm architecture is proposed in this paper. And also an improved single string matching algorithm based on a variant Boyer-Moore algorithm is presented. We implement our algorithm on the above architecture and the experiments prove that it is really practical and efficient on distributed memory machine. Its computation complexity is O(n/p + m), where n is the length of the text, and m is the length of the pattern, and p is the number of the processors.

Keywords: Boyer-Moore algorithm, distributed algorithm, parallel string matching, string matching.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2189

Authors: Yifan Fan, Xudong Luo, Pingping Lin

Abstract:

Nowadays, dialogue systems increasingly become the way for humans to access many computer systems. So, humans can interact with computers in natural language. A dialogue system consists of three parts: understanding what humans say in natural language, managing dialogue, and generating responses in natural language. In this paper, we survey deep learning based methods for dialogue management, response generation and dialogue evaluation. Specifically, these methods are based on neural network, long short-term memory network, deep reinforcement learning, pre-training and generative adversarial network. We compare these methods and point out the further research directions.

Keywords: Dialogue management, response generation, reinforcement learning, deep learning, evaluation.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 787

Authors: Motahar Reza, Anadi Sankar Gupta

Abstract:

An analysis is made of the flow of an incompressible viscoelastic fluid (of small memory) over a porous plate subject to suction or blowing. It is found that velocity at a point increases with increase in the elasticity in the fluid. It is also shown that wall shear stress depends only on suction and is also independent of the material of fluids. No steady solution for velocity distribution exists when there is blowing at the plate. Temperature distribution in the boundary layer is determined and it is found that temperature at a point decreases with increase in the elasticity in the fluid.

Keywords: Viscoelastic fluid, Flow past a porous plate, Heat transfer

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1335

Authors: Ruibiao J. Guo, Brad Cain, Pierre Meunier

Abstract:

The Integrated Performance Modelling Environment (IPME) is a powerful simulation engine for task simulation and performance analysis. However, it has no high level cognition such as memory and reasoning for complex simulation. This article introduces a knowledge representation and reasoning scheme that can accommodate uncertainty in simulations of military personnel with IPME. This approach demonstrates how advanced reasoning models that support similarity-based associative process, rule-based abstract process, multiple reasoning methods and real-time interaction can be integrated with conventional task network modelling to provide greater functionality and flexibility when modelling operator performance.

Keywords: Computer-Generated Forces, Human Behaviour Representation, IPME, Modelling and Simulation, Uncertainty Reasoning

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2117

Authors: Sujala.D.Shetty, S.Vadivel, Sakshi Vaghella

Abstract:

Data mining is the process of sifting through large volumes of data, analyzing data from different perspectives and summarizing it into useful information. One of the widely used desktop applications for data mining is the Weka tool which is nothing but a collection of machine learning algorithms implemented in Java and open sourced under the General Public License (GPL). A web service is a software system designed to support interoperable machine to machine interaction over a network using SOAP messages. Unlike a desktop application, a web service is easy to upgrade, deliver and access and does not occupy any memory on the system. Keeping in mind the advantages of a web service over a desktop application, in this paper we are demonstrating how this Java based desktop data mining application can be implemented as a web service to support data mining across the internet.

Keywords: desktop application, Weka mining, web service

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 4081

Authors: Khalid Elasnaoui, Brahim Aksasse, Mohammed Ouanan

Abstract:

In this paper, we are interested in the problem of finding similar images in a large database. For this purpose we propose a new algorithm based on a combination of the 2-D histogram intersection in the HSV space and statistical moments. The proposed histogram is based on a 3x3 window and not only on the intensity of the pixel. This approach overcome the drawback of the conventional 1-D histogram which is ignoring the spatial distribution of pixels in the image, while the statistical moments are used to escape the effects of the discretisation of the color space which is intrinsic to the use of histograms. We compare the performance of our new algorithm to various methods of the state of the art and we show that it has several advantages. It is fast, consumes little memory and requires no learning. To validate our results, we apply this algorithm to search for similar images in different image databases.

Keywords: 2-D histogram, Statistical moments, Indexing, Similarity distance, Histograms intersection.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1931

Authors: C. Kalamani, K. Paramasivam

Abstract:

In VLSI, testing plays an important role. Major problem in testing are test data volume and test power. The important solution to reduce test data volume and test time is test data compression. The Proposed technique combines the bit maskdictionary and 2n pattern run length-coding method and provides a substantial improvement in the compression efficiency without introducing any additional decompression penalty. This method has been implemented using Mat lab and HDL Language to reduce test data volume and memory requirements. This method is applied on various benchmark test sets and compared the results with other existing methods. The proposed technique can achieve a compression ratio up to 86%.

Keywords: Bit Mask dictionary, 2n pattern run length code, system-on-chip, SOC, test data compression.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1921

Authors: Sheena Mathew, K. Poulose Jacob

Abstract:

This paper describes the results of an extensive study and comparison of popular hash functions SHA-1, SHA-256, RIPEMD-160 and RIPEMD-320 with JERIM-320, a 320-bit hash function. The compression functions of hash functions like SHA-1 and SHA-256 are designed using serial successive iteration whereas those like RIPEMD-160 and RIPEMD-320 are designed using two parallel lines of message processing. JERIM-320 uses four parallel lines of message processing resulting in higher level of security than other hash functions at comparable speed and memory requirement. The performance evaluation of these methods has been done by using practical implementation and also by using step computation methods. JERIM-320 proves to be secure and ensures the integrity of messages at a higher degree. The focus of this work is to establish JERIM-320 as an alternative of the present day hash functions for the fast growing internet applications.

Keywords: Cryptography, Hash function, JERIM-320, Messageintegrity

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2641

Authors: Nicolas Zufferey

Abstract:

Ant algorithms are well-known metaheuristics which have been widely used since two decades. In most of the literature, an ant is a constructive heuristic able to build a solution from scratch. However, other types of ant algorithms have recently emerged: the discussion is thus not limited by the common framework of the constructive ant algorithms. Generally, at each generation of an ant algorithm, each ant builds a solution step by step by adding an element to it. Each choice is based on the greedy force (also called the visibility, the short term profit or the heuristic information) and the trail system (central memory which collects historical information of the search process). Usually, all the ants of the population have the same characteristics and behaviors. In contrast in this paper, a new type of ant metaheuristic is proposed, namely SMART (for Solution Methods with Ants Running by Types). It relies on the use of different population of ants, where each population has its own personality.

Keywords: Optimization, Metaheuristics, Ant Algorithms, Evolutionary Procedures, Population-Based Methods.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1720

Authors: Pornpimol Bungkomkhun, Surapong Auwatanamongkol

Abstract:

This paper presents a supervised clustering algorithm, namely Grid-Based Supervised Clustering (GBSC), which is able to identify clusters of any shapes and sizes without presuming any canonical form for data distribution. The GBSC needs no prespecified number of clusters, is insensitive to the order of the input data objects, and is capable of handling outliers. Built on the combination of grid-based clustering and density-based clustering, under the assistance of the downward closure property of density used in bottom-up subspace clustering, the GBSC can notably reduce its search space to avoid the memory confinement situation during its execution. On two-dimension synthetic datasets, the GBSC can identify clusters with different shapes and sizes correctly. The GBSC also outperforms other five supervised clustering algorithms when the experiments are performed on some UCI datasets.

Keywords: supervised clustering, grid-based clustering, subspace clustering

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1610

Authors: Seulbi Choi, Hyunchul Ahn

Abstract:

Collaborative filtering (CF) algorithm has been popularly used for recommender systems in both academic and practical applications. It basically generates recommendation results using users’ numeric ratings. However, the additional use of the information other than user ratings may lead to better accuracy of CF. Considering that a lot of people are likely to share their honest opinion on the items they purchased recently due to the advent of the Web 2.0, user's review can be regarded as the new informative source for identifying user's preference with accuracy. Under this background, this study presents a hybrid recommender system that fuses CF and user's review mining. Our system adopts conventional memory-based CF, but it is designed to use both user’s numeric ratings and his/her text reviews on the items when calculating similarities between users.

Keywords: Recommender system, collaborative filtering, text mining, review mining.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1587

Authors: Mihails Kulikovs, Ernests Petersons

Abstract:

Network management techniques have long been of interest to the networking research community. The queue size plays a critical role for the network performance. The adequate size of the queue maintains Quality of Service (QoS) requirements within limited network capacity for as many users as possible. The appropriate estimation of the queuing model parameters is crucial for both initial size estimation and during the process of resource allocation. The accurate resource allocation model for the management system increases the network utilization. The present paper demonstrates the results of empirical observation of memory allocation for packet-based services.

Keywords: Queuing System, Packet Loss Probability, Measurement-Based Admission Control (MBAC), Performanceevaluation, Quality of Service (QoS).

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1773

Authors: Wendy Zhang, Ernst L. Leiss, Huilin Ye

Abstract:

Most scientific programs have large input and output data sets that require out-of-core programming or use virtual memory management (VMM). Out-of-core programming is very error-prone and tedious; as a result, it is generally avoided. However, in many instance, VMM is not an effective approach because it often results in substantial performance reduction. In contrast, compiler driven I/O management will allow a program-s data sets to be retrieved in parts, called blocks or tiles. Comanche (COmpiler MANaged caCHE) is a compiler combined with a user level runtime system that can be used to replace standard VMM for out-of-core programs. We describe Comanche and demonstrate on a number of representative problems that it substantially out-performs VMM. Significantly our system does not require any special services from the operating system and does not require modification of the operating system kernel.

Keywords: I/O Management, Out-of-core, Compiler, Tile mapping.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1318