Parallel Client/Server System for Programming Language Research

计算机编程语言和运行时系统是编写顺序和并发程序的基本机制，将人类的想法转换为计算机可执行的实用的、功能强大的形式。所有的研究和工业软件都是从编程语言构建的，编程的趋势是将多个处理器(核心)组织为共享内存或分布在多台计算机上。使编程语言和运行时系统更具表现力、更易于使用、更安全且性能跨越所有计算学科。申请者(Buhr、Dietl、Karsten、Lam、Lhotak、Richards和Zhang)构建或改编这些基础工具，增强了所有计算机科学的基础设施，从教授编程到开发高级软件项目。购买新的实验性多核计算机(服务器)和测试硬件(交换机和客户端)对于支持滑铁卢大学的编程语言教师和研究生至关重要。该设备对于提供可靠的编程和实验环境至关重要，使申请者的研究在顺序和并行计算、性能测试和数据分析方面更上一层楼。如果没有下一代并行处理器，申请者就无法产生在实践中有意义或被认为与学术界相关的研究结果。该设备将用于开发和评估高级算法、软件技术和测试程序(语言设计、类型系统、并发性)和运行时系统(用户级线程、锁定、调度)的多个编程语言项目。该研究计划采取整体方法，在整个语言开发环境中进行垂直研究，即设计、编译、工具和高性能执行，特别是在并行计算机上。研究结果直接适用于现有的和新的编程语言，以及相关的高性能运行时系统。我们集团的毕业生精通复杂的系统级软件的开发和工作，在IT行业的高级水平上非常受欢迎。这项研究工作及其相关软件的潜在用户是加拿大IT行业的多个部门，从小型和创业型软件公司到设备供应商和电信供应商。这些软件公司将能够快速生产复杂的并发应用程序，有效地利用现在和未来可用的多处理器技术。反过来，这又通过降低开发成本和上市时间为这些公司创造了经济优势。此外，研究软件是开源的，因此加拿大公司和消费者不受专有和昂贵的系统和应用程序的束缚，这为企业和消费者节省了成本。

Computer programming languages and runtime systems are the foundational mechanism for writing sequential and concurrent programs, transforming human ideas into a practical, functioning form executable by a computer. All research and industrial software are built from programming languages, and programming is trending towards multiple processors (cores) organized as shared memory or distributed across multiple computers. Making programming languages and runtime systems more expressive, easier to use, safer, and performant cuts across all computing disciplines. The applicants (Buhr, Dietl, Karsten, Lam, Lhotak, Richards, and Zhang) build or adapt these foundational tools, enhancing the infrastructure for all of computer science, from teaching programming to developing advanced software projects. The acquisition of new experimental multi-core computers (servers) and testing hardware (switch and clients) is critical for supporting programming-language faculty and graduate students at the University of Waterloo. The equipment is essential to provide a reliable programming and experimental environment to move the applicants' research to the next level with respect to sequential and parallel computation, performance testing, and data analysis. Without access to the next generation of parallel processors, the applicants are unable to produce research results that matter in practice or are considered relevant in academia. The equipment will be used in multiple programming-language projects developing and evaluating advanced algorithms, software techniques, and testing procedures (language design, type systems, concurrency) and runtime systems (user-level threading, locking, scheduling). The research plan takes a holistic approach, performing research vertically across the entire language-development environment, i.e., design, compilation, tools, and high-performance execution especially on parallel computers. The research results are directly applicable to existing and new programming languages, and associated high performance runtime-systems. Graduates from our group are proficient at developing and working on sophisticated and complex system-level software, and are highly sought after at advanced levels in the IT industry. The potential users of this research work and its associated software are multiple sectors of the Canadian IT industry, ranging from small-scale and entrepreneurial software companies to equipment vendors and telecommunication providers. These software companies will be able to rapidly produce complex, concurrent applications, which effectively exploit multiprocessor technology available today and tomorrow. In turn, this creates an economic advantage for these companies by lowering development costs and time to market. As well, the research software is open-source so Canadian companies and consumers are not tied to proprietary and expensive systems and applications, which results in cost savings for businesses and consumers.