Supplement: CISE Research Infrastructure: Modeling Complex Physical and Computational Environments

9623614犹他州威廉大学cise研究基础设施：为复杂的物理和计算环境建模该奖项授予建模与仿真支持机构，该机构为在新型计算机体系结构、制造和设计、过程控制、远程操作和医学中的终端用户应用程序的工作构成了重要的基础设施；同时也是机器体系结构、硬件和软件系统创新的试验台。建模设施将使用最先进的技术来实现高带宽、低延迟的机器到机器通信，从而产生显著的计算优势。在紧耦合的通信环境中使用多个不同的和潜在的专用计算机允许模型构建者探索可以极大地减少模拟运行时间的新的并行模拟方法。这种能力的提高允许解决全新类别的关键问题。现代模拟产生了大量数据，对这些数据的解释可能不会立即显现出来。因此，建模设施的第二个中心组件是大容量存储管理系统。最后，建模设施将配备几个特殊用途的设备，这些设备可以利用通信和数据管理设施来解决目前在传统学术计算环境中不可行的大规模问题。

9623614 Thompson, William University of Utah CISE Research Infrastructure: Modeling Complex Physical and Computational Environment This award is for the creation of a Modeling and Simulation Support Facility, constituting vital infrastructure for work on end user applications in novel computer architectures, manufacturing and design, process control, teleoperations, and medicine; while also serving as a testbed for innovations in machine architecture, hardware, and software systems. The Modeling Facility will use state-of-the-art technology to achieve high-bandwidth, low-latency machine-to-machine communications, resulting in significant computational leverage. The use of multiple heterogeneous and potentially specialized computers in a tightly coupled communication environment permits model builders to explore novel parallel simulation methods that can greatly reduce simulation run times. This increase in capability permits entirely new classes of critical problems to be solved. Modern simulations produce large quantities of data, the interpretation of which may not be immediately apparent. Thus, the second central component of the Modeling Facility is a large capacity storage management system. Finally, the Modeling Facility will be equipped with several special purpose devices that can exploit the communications and data management facilities to tackle large-scale problems that are currently infeasible in a traditional academic computing environment.