PURCHASE OF SGI WORKSTATIONS AND MULTIPROCESSOR SERVER

要求支持一个新的工作站集群，以满足UCI化学系九名NIH资助的教员及其研究小组的计算需求。该集群由一台多处理器服务器和六台安装了分子建模软件的Silicon Graphics O2 R1200工作站组成。需要所需计算资源的由美国国立卫生研究院资助的研究项目有：1)生物聚合物和模型系统的结构建模：对多糖的结构和动力学进行建模；通过活性部位建模和文库搜索设计新的磷酸酶抑制剂；预测和理解人工β-折叠的结构；对肺表面活性物质蛋白质的结构和动力学进行建模。最后一位是尚未获得NIH支持的新员工。2)预测有机反应性和选择性：设计各种新的立体选择性反应；合成复杂的天然产物和其他生物和医药目标。这里描述的NIH支持的研究项目的数量和广度都很大。虽然大多数研究人员主要是实验者，但大量的计算工作是所有这些项目的关键组成部分。研究人员依靠集中式设施在大量研究人员中最有效地利用CPU和图形资源，这些研究人员主要是实验者，因此是断断续续的用户，并节省了成本。一台重要的多处理器计算机将允许进行目前我们无法访问的更大规模的计算。UCI强大的化学理论小组，分子建模设备经理，具有计算化学专业知识的博士化学家，并致力于培训建模方法方面的研究生研究人员，使我们现有的共享计算资源变得非常有价值。现有的硬件正在变得过时，新的硬件已成为迫切需要。UCI已经通过设施经理的工资为计算化学研究和指导方面的卓越做出了重要的长期资金承诺。

Support for a new workstation cluster is requested to meet the computing needs of nine NIH-funded faculty and their research groups in the Department of Chemistry at UCI. The cluster consists of a multi- processor server and six Silicon Graphics O2 R1200 workstations with molecular modeling software. NIH-supported research projects for which the requested computing resources is needed are: 1) Structural modeling of biopolymers and model systems: Modeling the structure and dynamics of polysaccharides; design of new phosphatase inhibitors through active-site modeling and library searching; predicting and understanding the structure of artificial beta-sheets; modeling structure and dynamics of pulmonary surfactant proteins. The last is a new hire who does not yet have NIH support. 2) Prediction of organic reactivity and selectivity: Design of a variety of new stereoselective reactions; synthesis of complex natural products and other targets of biological and pharmaceutical interest. The number and breadth of the NIH-supported research projects described herein are large. While most of the researchers are primarily experimentalists, significant computational work is a key component of all of these projects. Researchers rely on a centralized facility for most effective utilization of CPU and graphics resources among a large number of researchers who are primarily experimentalists and thus intermittent users, and for cost savings. A significant multi-processor computer will allow larger scale calculations currently inaccessible to us. The strong Chemistry theory group at UCI, a Molecular Modeling Facility Manger who is a Ph.D. chemist with computational chemistry expertise, and a commitment to training graduate researchers in modeling methods who have made our existing shared computing resources extremely valuable. Existing hardware is becoming obsolete and new hardware has become an urgent need. UCI has made a significant long-term funding commitment to excellence in computational chemistry research and instructions through the salary of the Facility Manager.