SPX: Collaborative Research: Moving Towards Secure and Massive Parallel Computing

SPX：协作研究：迈向安全和大规模并行计算

• 批准号: 1822809
• 负责人: Clifford Stein
• 金额: $ 6.83万
• 依托单位: Columbia University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2018
• 资助国家: 美国
• 起止时间: 2018-10-01 至 2020-09-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1822809&HistoricalAwards=false
• 关键词: SPX; Collaborative; Research; Moving; Towards

Modern computing systems have moved beyond single-core, single-processor devices to more modern multi-core parallel processors operating in networked systems and available in warehouse-scale clouds popularized by industries and the government. This new parallel, interconnected, big-data world requires fundamental research on multiple levels from algorithms to systems and computer architecture. This project seeks to take initial steps in the study of the expansive set of algorithms and systems issues in this important research challenge by building and developing new general frameworks for massive parallel computation, often involving privacy and security, in real-life scenarios. The investigators? long-term goals include two directions. As the first thrust of this effort, the investigators aim to design fundamental and efficient algorithms for massive parallel computations in the practical MapReduce framework, in particular by reducing the number of rounds in this framework. As the second thrust of this effort, the investigators aim to augment current parallel environments and architectures with better data structures and abstractions to develop simplified and fast implementations of fundamental algorithms such that everyone can use them in practice.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

现代计算系统已经超越单核、单处理器设备，转向在联网系统中操作的更现代的多核并行处理器，并且可用于由行业和政府推广的仓库规模的云。这个新的并行、互联的大数据世界需要从算法到系统和计算机架构的多个层面进行基础研究。该项目旨在通过构建和开发新的大规模并行计算的通用框架，在现实生活中，通常涉及隐私和安全，在这一重要的研究挑战中研究广泛的算法和系统问题。调查员？长期目标包括两个方向。 作为这项工作的第一个推力，研究人员的目标是设计基本的和有效的算法，在实际的MapReduce框架中的大规模并行计算，特别是通过减少在这个框架中的轮数。作为这项工作的第二个重点，研究人员的目标是用更好的数据结构和抽象来增强当前的并行环境和架构，以开发简化和快速的基本算法实现，以便每个人都可以在实践中使用它们。该奖项反映了NSF的法定使命，并被认为值得通过使用基金会的智力价值和更广泛的影响审查标准进行评估来支持。

项目成果

Distributed Algorithms for Matching in Hypergraphs

超图中的分布式匹配算法

• 发表时间: 2020
• 期刊: Workshop on Approximation and Online Algorithms
• 作者: Hanguir, O. Stein

Parallel approximate undirected shortest paths via low hop emulators

通过低跳模拟器并行近似无向最短路径

• DOI: 10.1145/3357713.3384321
• 发表时间: 2020
• 期刊: Symposium on Theory of Computing (STOC
• 作者: Andoni, Alexandr;Stein, Clifford;Zhong, Peilin
• 通讯作者: Zhong, Peilin

Coresets Meet EDCS: Algorithms for Matching and Vertex Cover on Massive Graphs

核心集满足 EDCS：海量图上的匹配和顶点覆盖算法

• 发表时间: 2019
• 期刊: SODA 2019
• 作者: Assadi, S. Batenai

Log Diameter Rounds Algorithms for 2-Vertex and 2-Edge Connectivity

用于 2 顶点和 2 边连接的对数直径舍入算法

• DOI: 10.4230/lipics.icalp.2019.14
• 发表时间: 2019
• 期刊: and Programming
• 作者: Andoni, Alexandr;Stein, Clifford;Zhong, Peilin
• 通讯作者: Zhong, Peilin