Hardware and Software Co-Design for High Performance Systems

高性能系统的硬件和软件协同设计

• 批准号: 9554561
• 负责人: Jerzy Rozenblit
• 金额: $ 56.25万
• 依托单位: University of Arizona
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 1995
• 资助国家: 美国
• 起止时间: 1995-09-15 至 2002-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=9554561&HistoricalAwards=false
• 关键词: Hardware; Software; Co; Design; Performance

GER-9554561 Rozenblit Hardware/Software Codesign is receiving increasingly widespread attention. Codesign provides methods and techniques which can effectively and efficiently manage the diversity and complexity of applications employing embedded systems. They result in a decreased development cost of complex systems and their time to market deployment. Our proposed research targets are: a)implementation independent specification and modeling of embedded systems, b) associated partitioning, based on iterative refinement, c) construction of modules for mixed software/hardware components, and d) component mapping to target implementations. These key target areas will form the foundation for the development of a graduate research training program. The baseline for the proposed effort is our internationally recognized research in the areas of codesign, computer systems engineering, design and synthesis of VLSI systems, and advanced work in multi-chip module (MCM) partitioning and behavioral synthesis. Technical Innovation: In the proposes research, concepts, techniques and tools for the design of high-performance, low-cost systems that consists of hardware and software components will be developed that transform an implementation independent, model- based description into a working prototype. Structural Innovation: Joint work and continuous interaction with local (State of Arizona) industrial partners, international collaborators, student exchanges, and industrial mentorship. The development of a state-of-the-art graduate program in codesign providing a broad based, academic and industrial experience and thus training future science and engineering leaders.

格尔-9554561 Rozenblit硬件/软件协同设计正受到越来越广泛的关注。协同设计提供了一些方法和技术，可以有效地管理采用嵌入式系统的应用程序的多样性和复杂性。它们降低了复杂系统的开发成本，缩短了市场部署时间。 我们提出的研究目标是：a）实现独立的规格说明和建模的嵌入式系统，B）相关的分区，基于迭代细化，c）建设的模块混合软件/硬件组件，和d）组件映射到目标实现。 这些关键目标领域将形成研究生研究培训计划发展的基础。 拟议工作的基线是我们在协同设计，计算机系统工程，VLSI系统的设计和综合领域的国际公认的研究，以及多芯片模块（MCM）划分和行为综合的先进工作。 技术创新：在提出的研究，概念，技术和工具的设计，高性能，低成本的系统，包括硬件和软件组件将开发转换为一个独立的实现，基于模型的描述到一个工作原型。 结构创新：与当地（亚利桑那州）工业合作伙伴，国际合作者，学生交流和工业导师的联合工作和持续互动。 开发最先进的协同设计研究生课程，提供广泛的学术和工业经验，从而培养未来的科学和工程领导者。