Polynomials are widely used in scientific computing and engineering. In this paper, we present an accurate and fast compensated algorithm to evaluate bivariate polynomials with floating-point coefficients. this algori...
详细信息
ISBN:
(纸本)9781509050819
Polynomials are widely used in scientific computing and engineering. In this paper, we present an accurate and fast compensated algorithm to evaluate bivariate polynomials with floating-point coefficients. this algorithm is applying error free transformations to the bivariate Horner scheme and sum the final decomposition accurately. We also prove the proposed algorithm's accuracy with forward error analysis that the accuracy of the computed result is similar to the result computed by the bivariate Horner scheme in twice the working precision. Numerical experiments illustrate the behavior and it has higher efficiency than the bivariate Horner scheme implemented in double-double library.
In this paper, a bit-rate estimation method is proposed, which could improve the performance of the H.264 encoder and avoid the strict data-dependences of mode decision. After that, an efficient parallel algorithm for...
详细信息
ISBN:
(纸本)9780769534435
In this paper, a bit-rate estimation method is proposed, which could improve the performance of the H.264 encoder and avoid the strict data-dependences of mode decision. After that, an efficient parallel algorithm for H.264 encoder with CABAC entropy coding is presented based on the Macro-Block Region Partition (MBRP) parallel method and the bit-rate estimation technique. Simulation results show that, the proposed parallel algorithm could improve the performance of H.264 encoder efficiently while maintaining the similar RD performance as JM 10.2.
D-ReServE increases reliability of SOA-based systems in case of failure occurrence. the fault-tolerant information in D-ReServE is stored in the Stable Storage, which available space depletes with time. thus, in this ...
详细信息
ISBN:
(纸本)9780769548791
D-ReServE increases reliability of SOA-based systems in case of failure occurrence. the fault-tolerant information in D-ReServE is stored in the Stable Storage, which available space depletes with time. thus, in this paper we propose a garbage collection protocol for D-ReServE that allows the periodic purging of the Stable Storage, and discuss the challenges of garbage collection due to the nature of SOA systems.
this paper proposes an approach to the design of distributedparallelcomputing systems based on the concept of reflective middleware, skeleton parallel programming and object-oriented software engineering. In the app...
详细信息
ISBN:
(纸本)0780378407
this paper proposes an approach to the design of distributedparallelcomputing systems based on the concept of reflective middleware, skeleton parallel programming and object-oriented software engineering. In the approach, the programmer does not need to be bothered by the architectural details. Furthermore, the approach provides support for the mostly used parallel paradigms, which will ease the onerous work for the programmer to construct the parallelism. It offers a method for the user to map the programming design seamlessly to the real environment.
We compare the performance of the Treadmarks DSM system with two popular message passing systems (PVM and MPI). the comparison is done on 1, 2, 4, 8, 16, 24, and 32 nodes. applications are chosen to represent three cl...
详细信息
ISBN:
(纸本)0780378407
We compare the performance of the Treadmarks DSM system with two popular message passing systems (PVM and MPI). the comparison is done on 1, 2, 4, 8, 16, 24, and 32 nodes. applications are chosen to represent three classes of problems: loosely synchronous, embarrassingly parallel, and synchronous. the results show DSM has similar performance to message passing for the embarrassingly parallel class. However, the performance of DSM is lower than PVM and MPI for the synchronous and loosely synchronous classes of problems. An analysis of the reasons is presented.
Designing fast singular value decomposition (SVD) is significantly interesting in applications. the random direct SVD (RSVD) has provided a fast scheme to compute the well approximate SVD by unilateral randomized samp...
详细信息
ISBN:
(纸本)9781509050819
Designing fast singular value decomposition (SVD) is significantly interesting in applications. the random direct SVD (RSVD) has provided a fast scheme to compute the well approximate SVD by unilateral randomized sampling. In this paper, we present an efficient random algorithm in a bilateral sampling way. We also prove that the proposed algorithms can be bounded well and have less computational complexity compared to RSVD when the objective matrix is approximately square. Numerical experiments on graph Laplacian and Hilbert matrix demonstrate the efficiency and stability of the proposed methods.
Edge computing extends a traditional cloud data centre model often by using a microdata centre (mdc) at the network edge for computation and storage. As these edge devices are in proximity to users, this results in im...
详细信息
ISBN:
(纸本)9781450351492
Edge computing extends a traditional cloud data centre model often by using a microdata centre (mdc) at the network edge for computation and storage. As these edge devices are in proximity to users, this results in improved application response times and reduces load on the cloud data center (cdc). In this paper, we propose a security and deadline aware scheduling algorithm called RT-SANE (Real-Time Security Aware scheduling on the Network Edge). applications with stringent privacy requirements are scheduled on an mdc closer to the user, whereas others can be scheduled on a cdc or a remotemdc. We also discuss how application performance and network latency influence the choice of an mdc or cdc. the intuition is that due to a lower communication latency between the user & the mdc, more applications are able to meet their deadlines when run on themdc. Conversely, applications with loose deadlines may be executed on a cdc. In order to facilitate this, we also propose a distributed orchestration architecture and protocol that is both performance & security aware. Simulation results show that RT-SANE offers superior real-time performance compared to a number of other scheduling policies in an Edge computing environment, while meeting application privacy requirements.
We describe a tool that implements a set of services to manipulate and store data from a radar network in a transparent way to end users. A major requirement of this system is data availability and reliability. Conseq...
详细信息
ISBN:
(纸本)9780769527369
We describe a tool that implements a set of services to manipulate and store data from a radar network in a transparent way to end users. A major requirement of this system is data availability and reliability. Consequently, we have implemented a redundancy schema based on the Information Dispersal Algorithm (IDA). Preliminary results show that the IDA based replication provides better reliability and less storage spending than traditional replication.
In a parallel processing system, a pancake graph is one of the superior topologies for interconnection network because of the small diameter and the high degree. In previous research, fault-tolerant routing using rest...
详细信息
ISBN:
(纸本)9781479983346
In a parallel processing system, a pancake graph is one of the superior topologies for interconnection network because of the small diameter and the high degree. In previous research, fault-tolerant routing using restricted global information called safety levels in a pancake graph was proposed. But there are some rooms for improvement. therefore, we propose a fault-tolerant routing algorithm based on improved safety levels to attain higher reachability. In addition, we estimated the proposed method in time and space complexities, and carried out a computer experiment to verify its effectiveness.
Energy consumption has increasingly become a serious problem in contemporary data centers. the electricity bill contributes a significant fraction of the Total Cost of Ownership (TCO), and it has been predicted to inc...
详细信息
ISBN:
(纸本)9780769548791
Energy consumption has increasingly become a serious problem in contemporary data centers. the electricity bill contributes a significant fraction of the Total Cost of Ownership (TCO), and it has been predicted to increase at an even faster pace in the following years. In an attempt to reduce the energy consumption in cloud infrastructures, we propose REST, an energy-efficient cloud storage built upon an open source cluster-based object store similar to GFS. It achieves high energy-efficiency by cleverly exploiting the redundancy and its asymmetric accesses existing in the system without compromising inherent well-established mechanisms, while maintaining reasonable performance level. By slightly altering the data-layout policy, REST can safely keep a large amount of redundant storage nodes in standby mode or even powered off most of the time. Deploying a sophisticated real-time workload monitor, it provides flexibility to power up standby or powered down nodes to accommodate the variations in workloads. In addition, trade-offs between energy and performance can be conveniently made by simply adjusting a trade-off metric in REST. the FileBench experimental results have demonstrated that power savings can be as high as 29%, still providing comparable or even better performance.
暂无评论