STTR Phase I: Ultra-large and low-cost electrodynamic modeling in commercial clouds

STTR第一阶段：商业云中的超大规模、低成本电动力学建模

基本信息

批准号：
1549673
负责人：
John Moore
金额：
$ 22.5万
依托单位：
FlexCompute Inc
依托单位国家：
美国
项目类别：
Standard Grant
财政年份：
2016
资助国家：
美国
起止时间：
2016-01-01 至 2017-06-30
项目状态：
已结题

来源：
https://www.nsf.gov/awardsearch/showAward?AWD_ID=1549673&HistoricalAwards=false
关键词：
STTR Phase Ultra large low

项目摘要

The broader impact/commercial potential of this Small Business Innovation Research Phase I project is to enable anyone with internet to access extremely powerful computing facilities to perform scientific computing. High performance computing is essential for computational prototyping in many areas of modern engineering, such as solar cells, automobiles, aerospace, and communication devices. However, significant barriers make it difficult for ordinary engineers to easily access powerful computing facilities. By leveraging the low-cost commercial cloud, the project will develop key technology that allow commercial clouds to perform extremely high performance scientific computing, thereby significantly reducing the cost and increasing the accessibility of high performance computing.This Small Business Innovation Research (SBIR) Phase I project will develop technologies that overcome key challenges in using commercial clouds for high performance computing. Unlike high performance computers, commercial clouds are not designed for scientific computing. They have high latency, intermitted availability, and heterogeneously distributed resources. These features make it difficult to conduct scientific computing. This project will develop new computing methods that fully take into account the heterogeneous nature of the commercial cloud. It enables anyone to access high performance scientific computing anywhere with an easy-to-use interface. Computing is provided as on-demand service. Engineers would not need to own any high performance computers and the computing resources are virtually unlimited.

这个小企业创新研究第一阶段项目更广泛的影响/商业潜力是使任何拥有互联网的人都能够访问极其强大的计算设施来执行科学计算。高性能计算对于现代工程许多领域（例如太阳能电池、汽车、航空航天和通信设备）的计算原型设计至关重要。然而，巨大的障碍使得普通工程师很难轻松获得强大的计算设施。该项目将利用低成本商业云，开发使商业云能够执行极高性能科学计算的关键技术，从而显着降低成本并提高高性能计算的可及性。本次小型企业创新研究（SBIR）阶段我的项目将开发技术来克服使用商业云进行高性能计算的关键挑战。与高性能计算机不同，商业云并不是为科学计算而设计的。它们具有高延迟、间歇性可用性和异构分布的资源。这些特征使得进行科学计算变得困难。该项目将开发新的计算方法，充分考虑商业云的异构性。它使任何人都可以通过易于使用的界面在任何地方访问高性能科学计算。计算作为按需服务提供。工程师不需要拥有任何高性能计算机，并且计算资源几乎是无限的。