CHEMICAL, BIOLOGICAL AND DRUG DISCOVERY COMPUTING SYSTEM

对于参与NIH支持的50多个项目的45个教员实验室的135名用户来说，计算在推动化学、生物和制药领域的知识方面发挥着核心作用。计算化学、分子相互作用研究和分子可视化是主要的应用领域。迫切需要一种最先进的计算硬件的分布式系统来支持这些项目的需求。这种需求因几个因素而变得更加突出。首先，计算包的多个冗余软件许可证已经合并，提供了额外的软件功能，但增加了用户数量，并使工作站容量变得紧张。其次，从多个服务器上消除冗余软件给单个临时服务器带来了更大的负载。第三，软件的升级使集中计算地点的一些功率较低的工作站过时。我们寻求通过购买(1)具有足够内存和磁盘存储资源以支持用户组需求的高性能多处理器服务器，以及(2)将位于现有中心站点的配备充分的工作站和图形输出设备来满足这一需求。系统支持将由13名经验丰富的集中工作人员提供，财务业务将通过提供共享计算设施超过12年的结构进行管理。拟议的硬件采购，加上用户支持的软件，将提供更大的处理能力、增强的可视化能力，并为开发可过渡到大学超级计算机的并行处理方法提供一个平台。它的可用性不仅将提高生产率，还将使大量NIH支持的用户可以使用一系列先进的软件。

For 135 users in 45 faculty laboratories engaged in more than 50 NIH- supported projects, computation plays a central role in advancing knowledge in the chemical, biological, and pharmaceutical fields. Computational chemistry, investigations of molecular interactions, and molecular visualization are major applications. There exists a pressing need for a distributed system of state-of-the-art computational hardware to support the needs of these projects. This need has been heightened by several factors. First, multiple redundant software licenses for computational packages have been coalesced, providing additional software capabilities but increasing the number of users and straining workstation capacity. Second, the elimination of redundant software from multiple servers has placed greater load on a single temporary server. Third, upgrading of the software has made obsolete some lower-powered workstations in centralized computational locations. We seek to address this need through the purchase of (1) a high-performance multi-processor server with adequate memory and disk storage resources to support the needs of the user group, and (2) adequately-equipped workstations and a graphics output device which will be located in an existing central site. System support will be provided by an experienced, centralized staff of 13, and financial operations will be managed through a structure that has provided for shared computational facilities for more than 12 years. The proposed hardware acquisition, coupled with user-supported software, would provide greater processing power, enhanced visualization capabilities, and a platform for development of parallel processing approaches that can be transitioned to the University's supercomputers. Its availability would not only enhance productivity, but also make a wide range of advanced software available to a significant number of NIH-supported users.