The proceedings contain 43 papers. The special focus in this conference is on algorithms andarchitectures for parallelprocessing. The topics include: CRFs for Digital Signature and NIZK Proof System in...
ISBN:
(纸本)9783031226762
The proceedings contain 43 papers. The special focus in this conference is on algorithms andarchitectures for parallelprocessing. The topics include: CRFs for Digital Signature and NIZK Proof System in Web Services;SPAC: Scalable Pattern Approximate Counting in Graph Mining;haica: A High Performance Computing & Artificial Intelligence Fused Computing Architecture;AOA: Adaptive Overclocking Algorithm on CPU-GPU Heterogeneous Platforms;GEM: Execution-Aware Cache Management for Graph Analytics;EnergyCIDN: Enhanced Energy-Aware Challenge-Based Collaborative Intrusion Detection in Internet of Things;Federated Learning-Based Intrusion Detection on Non-IID Data;long-Term Fairness Scheduler for Pay-as-You-Use Cache Sharing Systems;MatGraph: An Energy-Efficient and Flexible CGRA Engine for Matrix-Based Graph Analytics;pCOVID: A Privacy-Preserving COVID-19 Inference Framework;D-IOCost: Dynamic Cost-Aware Fair Queueing for Better I/O Proportionality and Performance;automated Binary Analysis: A Survey;LTNoT: Realizing the Trade-Offs Between Latency and Throughput in NVMe over TCP;AS-cast: Lock Down the Traffic of Decentralized Content Indexing at the Edge;Heterogeneous Graph Based Long- And Short-Term Preference Learning Model for Next POI Recommendation;SMTWM: Secure Multiple Types Wildcard Pattern Matching Protocol from Oblivious Transfer;a Label Flipping Attack on Machine Learning Model and Its Defense Mechanism;astute Approach to Handling Memory Layouts of Regular Data Structures;SparG: A Sparse GEMM Accelerator for Deep Learning Applications;An Efficient Transformer Inference Engine on DSP;hierarchical Reinforcement Learning-Based Mobility-Aware Content Caching and Delivery Policy for Vehicle Networks;GCNPart: Interference-Aware Resource Partitioning Framework with Graph Convolutional Neural Networks and Deep Reinforcement Learning;PipeFB: An Optimized Pipeline parallelism Scheme to Reduce the Peak Memory Usage;operator Placement for IoT Data Streaming Applications in Ed
Geometric modeling algorithms serve as the fundamental computation of CAD/CAM software in the field of computer graphics. The evaluation and derivative processes, being an essential component of geometric modeling alg...
详细信息
ISBN:
(纸本)9789819707973;9789819707980
Geometric modeling algorithms serve as the fundamental computation of CAD/CAM software in the field of computer graphics. The evaluation and derivative processes, being an essential component of geometric modeling algorithms, significantly impact their overall performance. However, when dealing with scenarios involving high-precision models or large-scale datasets, the lack of parallel acceleration for geometric modeling computation results in prolonged computation time and low computation efficiency, hindering the satisfactory experience of user interaction. Although the massive parallelism of GPUs has been proved with successful performance acceleration in various application fields, it has not been effectively utilized for accelerating geometric modeling algorithms. In this paper, we propose gGMED, a GPU-based approach specifically designed for accelerating the evaluation and derivative processes in geometric modeling. To leverage the massive parallel capability of GPU, our approach provides several optimizations such as data reuse, bank conflict avoidance, and pipeline execution, for effectively improving the performance of evaluation and derivative processes. The experiment results on representative GPUs and various NURBS models demonstrate that our approach can achieve up to 10.18x and34.56x performance speedup in end-to-end process and kernel computation respectively, compared to the state-of-the-art geometric modeling libraries.
The Random Ray Method (TRRM) is a new approach to solving partial differential equations (PDEs) based on the method of characteristics (MOC). It employs stochastic rather than deterministic discretization of character...
详细信息
ISBN:
(纸本)9789819708079;9789819708086
The Random Ray Method (TRRM) is a new approach to solving partial differential equations (PDEs) based on the method of characteristics (MOC). It employs stochastic rather than deterministic discretization of characteristic tracks and can be used for the numerical simulation of nuclear reactors. In this paper, we propose SW-TRRM, a parallel optimization program for TRRM based on the Sunway Bluelight II Supercomputer for the first time. We present a two-level parallelization scheme that consists of thread-level and process-level optimization. At the thread-level, we introduce three schemes for speeding up within a single core group, including direct parallelization, parallelization by energy groups, and loop structure optimization. At the process-level, we implement task parallelization among multiple processes using domain replication. Moreover, we devise an algorithm to optimize the MPI collective communication across super-nodes. Experimental results show that SW-TRRM achieves a 17.40x speedup within a single core group compared to the original TRRM program. When scaled up to 2,048 processes and 133,120 cores, SW-TRRM maintains good strong and weak scalability.
As blockchain technology garners increased adoption, permissioned blockchains like Hyperledger Fabric emerge as a popular blockchain system for developing scalable decentralized applications. Nonetheless, parallel exe...
详细信息
ISBN:
(数字)9789819708628
ISBN:
(纸本)9789819708611;9789819708628
As blockchain technology garners increased adoption, permissioned blockchains like Hyperledger Fabric emerge as a popular blockchain system for developing scalable decentralized applications. Nonetheless, parallel execution in Fabric leads to concurrent conflicting transactions attempting to read and write the same key in the ledger simultaneously. Such conflicts necessitate the abortion of transactions, thereby impacting performance. The mainstream solution involves constructing a conflict graph to reorder the transactions, thereby reducing the abort rate. However, it experiences considerable overhead during scenarios with a large volume of transactions or high data contention due to capture dependencies between each transaction. Therefore, one critical problem is how to efficiently order conflicting transactions during the ordering phase. In this paper, we introduce an optimized reordering algorithm designed for efficient concurrency control. Initially, we leverage key dependency instead of transaction dependency to build a conflict graph that considers read/write units as vertices and intra-transaction dependency as edges. Subsequently, a key sorting algorithm generates a serializable transaction order for validation. Our empirical results indicate that the proposed key-based reordering method diminishes transaction latency by 36.3% and considerably reduces system memory costs while maintaining a low abort rate compared to benchmark methods.
K-Means algorithm is one of the most common clustering algorithms widely applied in various data analysis applications. Yinyang K-Means algorithm is a popular enhanced K-Means algorithm that avoids most unnecessary ca...
详细信息
The emergence of mobile edge computing (MEC) has improved the data processing capabilities of devices with limited computing resources. However, some tasks that require higher latency and energy consumption are still ...
详细信息
ISBN:
(纸本)9783030953881;9783030953874
The emergence of mobile edge computing (MEC) has improved the data processing capabilities of devices with limited computing resources. However, some tasks that require higher latency and energy consumption are still facing huge challenges. In this paper, for the time-varying wireless channel conditions, we proposed an effective method to perform offloading calculations on the computing tasks of wireless devices, that is, to distribute the tasks to the local of offload to the edge server under the premise of satisfying time delay and energy consumption. Based on this, we adopt the parallel calculation model of Deep Reinforcement Learning Optimal Stopping Theory (DRLOST), which is composed of two parts: offloading decision generation and deep reinforcement learning. The model uses a parallel deep neural network (DNN) to generate offloading decisions, and stores the generated offloading decisions in the memory according to the optimal stopping theory model parameters to further train the model. The simulation results show that the proposed algorithm can minimize delay time, and can respond quickly to tasks even in a fast-fading environment.
Bandit Convex Optimization (BCO) is an imperative analysis framework when dealing with sequential decision-making problems. Considering to balance the computational cost and bounds of regrets, in this paper, we propos...
详细信息
With the increasing concern for environmental protection and resource optimization, efficient waste sorting has become a serious challenge today. In this paper, we propose a new offloading control problem that aims to...
详细信息
This book constitutes the refereed proceedings of the 22;internationalconference on algorithms andarchitectures for parallelprocessing, ica3pp 2022, which was held in October 2022. Due to COVID-19 pandemic the conf...
详细信息
ISBN:
(数字)9783031226779
ISBN:
(纸本)9783031226762
This book constitutes the refereed proceedings of the 22;internationalconference on algorithms andarchitectures for parallelprocessing, ica3pp 2022, which was held in October 2022. Due to COVID-19 pandemic the conference was held virtually.;The 33 full papers and 10 short papers, presented were carefully reviewed and selected from 91 submissions.;The papers cover many dimensions of parallelalgorithms andarchitectures, encompassing fundamental theoretical approaches, practical experimental projects, and commercial components and systems
暂无评论