CyberTraining: CIC: Widening the CI Workforce On-ramp by Exposing Undergraduates to Heterogeneous Computing

网络培训：CIC：通过让本科生接触异构计算来拓宽 CI 劳动力入口

• 批准号: 1829644
• 负责人: Apan Qasem
• 金额: $ 22.58万
• 依托单位: Texas State University - San Marcos
• 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: 2018-09-01 至 2023-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1829644&HistoricalAwards=false
• 关键词: CyberTraining; CIC; Widening; CI; Workforce

In keeping with NSF's mission of promoting scientific progress, this project strengthens the workforce of future cyberinfrastructure researchers and professionals by preparing undergraduate students to program and employ heterogeneous computing systems. The need for increased performance per watt and demands of processing diverse workloads have triggered a major industry shift towards systems containing different kinds of specialized components. These heterogeneous architectures are quickly becoming the dominant platform in high-performance computing (HPC), cloud computing, and Internet of Things (IoT) networks. As such, it is imperative that the scientific workforce in advanced cyberinfrastructure have a deep understanding of heterogeneity in computing systems. Current undergraduate computer science curricula lack sufficient coverage of heterogeneous computing concepts and there is an imminent risk that tomorrow's scientific workforce will be ill-equipped to program the complex heterogeneous systems of the future. This project addresses this by developing and disseminating modules covering heterogeneous computing concepts. Module development is complemented with out-of-class training camps that will serve as a springboard for undergraduates to pursue further training and career opportunities. The project takes an early-and-often, module-driven approach to curricular integration. A collection of modules is being designed to cover a range of heterogeneous computing concepts including heterogeneous architectures, hybrid algorithms, and heterogeneous programming models such as CUDA and OpenCL. For easy adoption, modules are designed to be self-contained with all requisite teaching material including in-class interactive exercises, problem sets and solutions, and pedagogical notes. Each module is designed to exploit heterogeneous context within the curriculum and introduce topics at an appropriate level of abstraction. The summer training camps are designed to be immersive experiences for undergraduates that (i) reinforce and extend the material covered in the modules, (ii) present the role of cyberinfrastructure in advancing scientific research, and (iii) expose students to opportunities in the fields of computational and data science. A key component of this project is forging alliances with industry partners, who are engaged throughout the design process. A detailed evaluation plan is in place to evaluate student learning outcomes and engagement. Both internal and external evaluators are involved in carrying out the evaluation plan. The modules and training activities are expected to impact over 1,100 computer science undergraduates enrolled at Texas State, Knox College, and Concordia University Texas. Extensive dissemination efforts through faculty training workshops are expected to further broaden the impact of this work.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.

为了与NSF促进科学进步的使命保持一致，该项目通过培养本科生编程和使用异构计算系统来加强未来网络基础设施研究人员和专业人员的队伍。对提高性能功耗比的需求和处理不同工作负载的需求引发了行业向包含不同类型专用组件的系统的重大转变。这些异构架构正迅速成为高性能计算（HPC）、云计算和物联网（IoT）网络中的主导平台。因此，先进网络基础设施中的科学工作者必须深刻理解计算系统的异质性。目前的本科计算机科学课程缺乏足够的覆盖面异构计算的概念，有一个迫在眉睫的风险，明天的科学劳动力将装备不良，以规划未来的复杂异构系统。 该项目通过开发和传播涵盖异构计算概念的模块来解决这一问题。模块开发与课外训练营相辅相成，这将成为本科生追求进一步培训和就业机会的跳板。 该项目采取早期和经常，模块驱动的方法来整合课程。正在设计一系列模块，以涵盖一系列异构计算概念，包括异构体系结构，混合算法和异构编程模型，如CUDA和OpenCL。为了便于采用，模块的设计是独立的所有必要的教学材料，包括课堂互动练习，问题集和解决方案，以及教学笔记。 每个模块的目的是利用课程中的异构环境，并在适当的抽象层次上介绍主题。夏令营旨在为本科生提供身临其境的体验，（i）加强和扩展模块中涵盖的材料，（ii）展示网络基础设施在推进科学研究中的作用，以及（iii）让学生接触计算和数据科学领域的机会。 该项目的一个关键组成部分是与参与整个设计过程的行业合作伙伴建立联盟。 我们制定了详细的评估计划，以评估学生的学习成果和参与度。 内部和外部评价人员都参与执行评价计划。 这些模块和培训活动预计将影响德克萨斯州立大学、诺克斯学院和德克萨斯州康考迪亚大学的1,100多名计算机科学本科生。 通过教师培训讲习班的广泛传播努力，预计将进一步扩大这项工作的影响。这个奖项反映了NSF的法定使命，并已被认为是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的支持。

项目成果

Accelerating HotSpots in Deep Neural Networks on a CAPI-Based FPGA

• DOI: 10.1109/hpcc/smartcity/dss.2019.00048
• 发表时间: 2019-08
• 期刊: 2019 IEEE 21st International Conference on High Performance Computing and Communications; IEEE 17th International Conference on Smart City; IEEE 5th International Conference on Data Science and Systems (HPCC/SmartCity/DSS)
• 作者: Md. Syadus Sefat;S. Aslan;J. W. Kellington;Apan Qasem

YODA: A Pedagogical Tool for Teaching Systems Concepts

YODA：用于教授系统概念的教学工具

• DOI: 10.1145/3478431.3499322
• 发表时间: 2022
• 期刊: Association for Computing Machinery
• 作者: Qasem, Apan

Characterizing Input-sensitivity in Tightly-Coupled Collaborative Graph Algorithms

表征紧耦合协作图算法中的输入敏感性

• DOI: 10.1109/ccgrid51090.2021.00038
• 发表时间: 2021
• 期刊: Cloud and Internet Computing (CCGrid
• 作者: Hope, Jacob M.;Gjergji, Mikel;Di Girolamo, Johana;Alvarez, Marco;Qasem, Apan
• 通讯作者: Qasem, Apan

Heterogeneous Computing for Undergraduates: Introducing the ToUCH Module Repository

本科生异构计算：ToUCH 模块存储库简介

• DOI: 10.1145/3478432.3499152
• 发表时间: 2022
• 期刊: Association for Computing Machinery
• 作者: Qasem, Apan;Bunde, David
• 通讯作者: Bunde, David

Short modules for introducing heterogeneous computing: conference tutorial

介绍异构计算的简短模块：会议教程

• DOI: 10.5555/3469567.3469575
• 发表时间: 2021
• 期刊: Journal of computing sciences in colleges
• 作者: Bunde, David P.;Qasem, Apan;Schielke, Philip
• 通讯作者: Schielke, Philip