检索结果-内蒙古大学图书馆

GPABE: GPU-Based Parallelization Framework for Attribute-Based Encryption Schemes

IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS 2025年第3期36卷 520-536页

作者： Xu, Wenhan Ma, Hui Zhang, Rui Li, Jianhao Chinese Acad Sci Inst Informat Engn IIE State Key Lab Cyberspace Secur Def Beijing 100093 Peoples R China Univ Chinese Acad Sci Sch Cybersecur Beijing 100049 Peoples R China

Attribute-based encryption (ABE) has emerged as a new paradigm for access control in cloud computing. However, despite the many promising features of ABE, its deployment in real-world systems is still limited, partially due to the expensive cost of its underlying mathematical operations, which often grow linearly with the size and complexity of the system's security policies. This becomes particularly challenging in data-intensive applications, where multiple users may simultaneously access and manipulate large volumes of data, resulting in high levels of con- currency and demand for computing resources, which are too heavy even for high-end servers. Further exacerbating the issues are the functionality and security requirements of a cloud, as they introduce additional computations to both the client and the server. Therefore, in this work, we introduce GPABE, the first GPU-based parallelization framework for ABE to facilitate its batch processing in cloud computing. By analyzing ABE's major computational workload, we identify multiple arithmetic modules that are common in the design of pairing-based ABEs. Based on the analysis, we further propose to decompose the ABE algorithm into computation graph, which can be efficiently implemented on the GPU platform. Our graph representation bridges the gap between ABE's high-level design and their low-level implementation on GPUs, and is applicable to a variety of popular schemes in the realm of ABE. We then implement GPABE as a heterogeneous computing server, with several optimization techniques to improve its throughput. Finally, we evaluate the GPU implementation of several ABE schemes using GPABE. The results show a speedup of least 51.0x and at most 253.6x for the throughput of ABE algorithms, compared to their state-of-the-art CPU implementations, which preliminarily demonstrated the effectiveness of GPABE.

关键词： graphics processing units Parallel processing Servers Encryption Optimization Arithmetic Security Instruction sets Costs Cloud computing Attribute-based encryption cloud computing data encryption parallel processing computation graph GPU

来源：评论

学校读者我要写书评

暂无评论

Edgeless-GNN: Unsupervised Representation Learning for Edgeless Nodes

引用

IEEE TRANSACTIONS ON EMERGING TOPICS IN COMPUTING 2024年第1期12卷 150-162页

作者： Shin, Yong-Min Tran, Cong Shin, Won-Yong Cao, Xin Yonsei Univ Sch Math & Comp Computat Sci & Engn Seoul 03722 South Korea Yonsei Univ Machine Intelligence & Data Sci Lab Seoul 03722 South Korea Posts & Telecommun Inst Technol Fac Informat Technol Hanoi 100000 Vietnam Univ New South Wales Sch Comp Sci & Engn Sydney 2052 Australia

We study the problem of embedding edgeless nodes such as users who newly enter the underlying network, while using graph neural networks (GNNs) widely studied for effective representation learning of graphs. Our study is motivated by the fact that GNNs cannot be straightforwardly adopted for our problem since message passing to such edgeless nodes having no connections is impossible. To tackle this challenge, we propose $\mathsf{Edgeless-GNN}$Edgeless-GNN, a novel inductive framework that enables GNNs to generate node embeddings even for edgeless nodes through unsupervised learning. Specifically, we start by constructing a proxy graph based on the similarity of node attributes as the GNN's computation graph defined by the underlying network. The known network structure is used to train model parameters, whereas a topology-aware loss function is established such that our model judiciously learns the network structure by encoding positive, negative, and second-order relations between nodes. For the edgeless nodes, we inductively infer embeddings by expanding the computation graph. By evaluating the performance of various downstream machine learning tasks, we empirically demonstrate that $\mathsf{Edgeless-GNN}$Edgeless-GNN exhibits (a) superiority over state-of-the-art inductive network embedding methods for edgeless nodes, (b) effectiveness of our topology-aware loss function, (c) robustness to incomplete node attributes, and (d) a linear scaling with the graph size.

关键词： computation graph edgeless node graph neural network (GNN) inductive network embedding unsupervised learning

来源：评论

学校读者我要写书评

暂无评论

A high-performance dataflow-centric optimization framework for deep learning inference on the edge

引用

JOURNAL OF SYSTEMS ARCHITECTURE 2024年 152卷

作者： Zhang, Runhua Jiang, Hongxu Geng, Jinkun Tian, Fangzheng Ma, Yuhang Wang, Haojie Beihang Univ Beijing Peoples R China Stanford Univ Stanford CA USA Tsinghua Univ Beijing Peoples R China

Edge computing has been emerging as a popular scenario for model inference. However, the inference performance on edge devices (e.g., Multi-Core DSP, FGPA, etc.) suffers from inefficiency due to the lack of highly optimized inference frameworks. Previous model inference frameworks are mainly developed in an operator-centric way, which provides insufficient acceleration to edge-based inference. Besides, the operator-centric framework incurs significant costs for continuous development and maintenance. Targeting the existing drawbacks of operator-centric frameworks, we design Xenos, which can automatically conduct dataflow-centric optimization of the computation graph and accelerate inference in two dimensions. Vertically, Xenos develops operator linking technique to improve data locality by restructuring the inter-operator dataflow. Horizontally, Xenos develops DSP-aware operator split technique to enable higher parallelism across multiple DSP units. Our evaluation demonstrates the effectiveness of vertical and horizontal dataflow optimization, which reduce the inference time by 15.0%-84.9% and 17.9%-89.9% , respectively. Besides, Xenos also outperforms the widely-used TVM by 1.1x-1.9x. Moreover, we extend Xenos to a distributed solution, which we call d-Xenos. d-Xenos employs multiple edge devices to jointly conduct the inference task and achieves a speedup of 3.68x-3.78x compared with the single device.

关键词： Edge computing Model inference Dataflow-centric computation graph Data locality

来源：评论

学校读者我要写书评

暂无评论

DLGR: A Rule-Based Approach to graph Replacement for Deep Learning 26

DLGR: A Rule-Based Approach to Graph Replacement for Deep Le...

引用

26th International Conference on Engineering of Complex Computer Systems (ICECCS)

作者： Ma, Enze Beijing Forestry Univ Dept Comp Sci Beijing Peoples R China

ISBN: (纸本)9781665401623

In deep learning libraries like TensorFlow, computations are manually batched as computation graphs. graph replacement is then an optimization that replaces one subgraph of a computation graph with another whilst keeping the graphs before and after replacement functionally equivalent. Meanwhile, in practice, it remains a challenge how graph replacements can be performed efficiently: graph replacement is usually conducted by human engineers, and thus it incurs many human efforts since a variety of deep learning models do exist and a number of model-specific replacements can be performed;the functionality equivalence of graphs before and after replacement is also not easy to guarantee. To tackle with this challenge, we introduce in this paper DLGR, a rule-based approach to graph replacement for deep learning. The core idea of DLGR is to define a set of replacement rules, each of which specifies the source and the target graph patterns and constraints on graph replacement. Given a computation graph, DLGR then performs an iterative process of matching and replacing subgraphs in the source graph, and generates a replaced, and usually optimized computation graph. We conduct experiments to evaluate the capabilities of DLGR. The results clearly show the strengths of DLGR: compared with two existing graph replacement techniques, it provides with more replacement rules and saves engineers' development efforts in reducing up to 68% lines of code.

关键词： computation graph graph replacement Replacement rule

来源：评论

学校读者我要写书评

暂无评论

GLite: A Fast and Efficient Automatic graph-Level Optimizer for Large-Scale DNNs 22

GLite: A Fast and Efficient Automatic Graph-Level Optimizer ...

引用

59th ACM/IEEE Design Automation Conference (DAC) - From Chips to Systems - Learn Today, Create Tomorrow

作者： Li, Jiaqi Peng, Min Li, Qingan Peng, Meizheng Yuan, Mengting Wuhan Univ Wuhan Peoples R China

ISBN: (纸本)9781450391429

We propose a scalable graph-level optimizer named GLite to speed up search-based optimizations on large neural networks. GLite leverages a potential-based partitioning strategy to partition large computation graphs into small subgraphs without losing profitable substitution patterns. To avoid redundant subgraph matching, we propose a dynamic programming algorithm to reuse explored matching patterns. The experimental results show that GLite reduces the running time of search-based optimizations from hours to milliseconds, without compromising in inference performance.

关键词： neural networks optimization computation graph inference

来源：评论

学校读者我要写书评

暂无评论

Enabling Resource-Aware Mapping of Spiking Neural Networks via Spatial Decomposition

引用

IEEE EMBEDDED SYSTEMS LETTERS 2021年第3期13卷 142-145页

作者： Balaji, Adarsha Song, Shihao Das, Anup Krichmar, Jeffrey Dutt, Nikil Shackleford, James Kandasamy, Nagarajan Catthoor, Francky Drexel Univ Dept Elect & Comp Engn Philadelphia PA 19104 USA Univ Calif Irvine Dept Comp Sci Irvine CA 92697 USA IMEC B-3001 Leuven Belgium Katholieke Univ Leuven ESAT B-3000 Leuven Belgium

With growing model complexity, mapping spiking neural network (SNN)-based applications to tile-based neuromorphic hardware is becoming increasingly challenging. This is because the synaptic storage resources on a tile, viz., a crossbar, can accommodate only a fixed number of presynaptic connections per postsynaptic neuron. For complex SNN models that have many presynaptic connections per neuron, some connections may need to be pruned after training to fit onto the tile resources, leading to a loss in the model quality, e.g., accuracy. In this letter, we propose a novel unrolling technique that decomposes a neuron function with many presynaptic connections into a sequence of homogeneous neural units, where each neural unit is a function computation node, with two presynaptic connections. This spatial decomposition technique significantly improves crossbar utilization and retains all presynaptic connections, resulting in no loss of the model quality derived from connection pruning. We integrate the proposed technique within an existing SNN mapping framework and evaluate it using machine learning applications on the DYNAP-SE state-of-the-art neuromorphic hardware. Our results demonstrate an average 60% lower crossbar requirement, $9\times $ higher synapse utilization, 62% lower wasted energy on the hardware, and between 0.8% and 4.6% increase in the model quality.

关键词： computation graph machine learning neuromorphic computing spiking neural networks (SNNs)

来源：评论

学校读者我要写书评

暂无评论

HARP: Holistic Analysis for Refactoring Python-Based Analytics Programs 20

HARP: Holistic Analysis for Refactoring Python-Based Analyti...

引用

42nd ACM/IEEE International Conference on Software Engineering - Companion Proceedings (ICSE-Companion) / 42nd ACM/IEEE International Conference on Software Engineering - Software Engineering in Practice (ICSE-SEIP)

作者： Zhou, Weijie Zhao, Yue Zhang, Guoqiang Shen, Xipeng North Carolina State Univ Raleigh NC 27695 USA Facebook Menlo Pk CA USA

ISBN: (纸本)9781450371216

Modern machine learning programs are often written in Python, with the main computations specified through calls to some highly optimized libraries (e.g., TensorFlow, PyTorch). How to maximize the computing efficiency of such programs is essential for many application domains, which has drawn lots of recent attention. This work points out a common limitation in existing efforts: they focus their views only on the static computation graphs specified by library APIs, but leave the influence from the hosting Python code largely unconsidered. The limitation often causes them to miss the big picture and hence many important optimization opportunities. This work proposes a new approach named HARP to address the problem. HARP enables holistic analysis that spans across computation graphs and their hosting Python code. HARP achieves it through a set of novel techniques: analytics-conscious speculative analysis to circumvent Python complexities, a unified representation augmented computation graphs to capture all dimensions of knowledge related with the holistic analysis, and conditioned feedback mechanism to allow risk-controlled aggressive analysis. Refactoring based on HARP gives 1.3-3X and 2.07X average speedups on a set of TensorFlow and PyTorch programs.

关键词： machine learning program computation graph dynamic language program analysis

来源：评论

学校读者我要写书评

暂无评论

Compiling Spiking Neural Networks to Mitigate Neuromorphic Hardware Constraints 11

Compiling Spiking Neural Networks to Mitigate Neuromorphic H...

引用

11th International Green and Sustainable Computing Workshop (IGSC)

作者： Balaji, Adarsha Das, Anup Drexel Univ Dept Elect & Comp Engn Philadelphia PA 19104 USA

ISBN: (纸本)9781665415521

Spiking Neural Networks (SNNs) are efficient computation models to perform spatio-temporal pattern recognition on resource- and power-constrained platforms. SNNs executed on neuromorphic hardware can further reduce energy consumption of these platforms. With increasing model size and complexity, mapping SNN-based applications to tile-based neuromorphic hardware is becoming increasingly challenging. This is attributed to the limitations of neuro-synaptic cores, viz. a crossbar, to accommodate only a fixed number of pre-synaptic connections per post-synaptic neuron. For complex SNN-based models that have many neurons and pre-synaptic connections per neuron, (1) connections may need to be pruned after training to fit onto the crossbar resources, leading to a loss in model quality, e.g., accuracy, and (2) the neurons and synapses need to be partitioned and placed on the neuro-sypatic cores of the hardware, which could lead to increased latency and energy consumption. In this work, we propose (1) a novel unrolling technique that decomposes a neuron function with many pre-synaptic connections into a sequence of homogeneous neural units to significantly improve the crossbar utilization and retain all pre-synaptic connections, and (2) SpiNeMap, a novel methodology to map SNNs on neuromorphic hardware with an aim to minimize energy consumption and spike latency.

关键词： Neuromorphic Computing Spiking Neural Networks (SNNs) Machine Learning computation graph

来源：评论

学校读者我要写书评

暂无评论

Model-checking task-parallel programs for data-race

引用

INNOVATIONS IN SYSTEMS AND SOFTWARE ENGINEERING 2019年第3-4期15卷 289-306页

作者： Nakade, Radha Mercer, Eric Aldous, Peter Storey, Kyle Ogles, Benjamin Hooker, Joshua Powell, Sheridan Jacob McCarthy, Jay Brigham Young Univ Provo UT 84602 USA Univ Massachusetts Lowell MA USA

Many of the correctness properties afforded by task-parallel programming models such as OpenMP, Cilk, X10, Chapel, Habanero, etc. rely on data-race freedom. The research in this paper studies data-race in the context of these models with the intent to prove with model checking its absence on any feasible schedule for a given input. The paper presents the computation graph as a representation of a happens-before relation that additionally tracks memory accesses with a quadratic algorithm to detect data-race on the graph. It then shows how the graph is constructed from an execution of a task-parallel program and proves that under a fixed order of mutual exclusion, if a schedule with a data-race exists in the program, then a data-race is manifest in the computation graph. The paper then defines a model checking algorithm that enumerates all orders of mutual exclusion to prove data-race freedom over all schedules on the given input. The approach is evaluated in a Java implementation of Habanero using the JavaPathfinder model checker. The results, when compared to other data-race detectors including one based on vector clocks, show that this new approach is more efficient than existing JavaPathfinder solutions and is comparable to the vector clock solution in the absence of data-race but slower in the presence of data-race since the vector clock algorithm is on-the-fly while the new approach is not. The results also show that the new approach avoids the memory overhead of vector clocks when there are many tasks and objects to track.

关键词： Model checking Data-race Task parallel Cilk x10 Habanero OpenMP SP-bags Vector clocks Formal verification Happens-before Partial order computation graph Determinism

来源：评论

学校读者我要写书评

暂无评论

An Incremental Approach to Scope-Bounded Checking Using a Lightweight Formal Method

An Incremental Approach to Scope-Bounded Checking Using a Li...

引用

2nd World Congress on Formal Methods/16th International Symposium on Formal Methods (FM 2009)

作者： Shao, Danhua Khurshid, Sarfraz Perry, Dewayne E. Univ Texas Austin Dept Elect & Comp Engn Austin TX 78712 USA

ISBN: (纸本)9783642050886

We present a novel approach to optimize scope-bounded checking programs using a relational constraint solver. Given a program and its correctness specification, the traditional approach translates a bounded code segment of the entire program into a declarative formula and uses a constraint solver to search for any correctness violations. Scalability is a key issue with such approaches since for non-trivial programs the formulas are complex and represent a heavy workload that can choke the solvers. Our insight is that hounded code segments, which can he viewed as a set of (possible) execution paths. naturally lend to incremental checking through a partitioning of the set, where each partition represents a sub-set of paths. The partitions can be checked independently, and thus the problem of scope-bounded checking for the given program reduces to several sub-problems. where each sub-problem requires the constraint solver to check a less complex formula, thereby likely reducing the solver's overall workload. Experimental results show that our approach provides significant speed-ups over the traditional approach.

关键词： Scope-bounded checking Alloy first-order logic SAT lightweight formal method computation graph white-box testing

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：