Equipment: MRI: Track 1 Acquisition of Current Hardware to Enhance Computational Research on the ELSA High Performance Computing Cluster at The College of New Jersey

在这个项目中，新泽西学院(TCNJ)将购买设备，以显著升级和增强我们的科学与分析电子实验室(ELSA)高性能计算(HPC)集群。作为一所以本科生为主的机构，TCNJ因吸引本科生参与研究而在全国范围内得到认可。在整个学年和夏季的本科生暑期体验(MUSE)中，理学院的教职员工与他们实验室的本科生密切合作，超过75%的理科毕业生获得了至少一学期的研究经验。在这个项目的过程中，每年有近100名本科生研究人员将受益于所有理学院系(物理、化学、生物、数学和统计、计算机科学)以及土木工程的实验室中由教师指导的研究。这些研究经验对学生来说是变革性的，使他们能够接受计算方面的培训，并促进他们在机器学习、天体物理、生物物理学、数学生物学和生物信息学等领域的研究。在研究实验室之外，理科教师将ELSA纳入他们的教学，每年让800-1000名本科生接触先进的计算。TCNJ还致力于为所有STEM学生创造一个研究密集型环境，通过提高代表性不足的学生和那些有较高经济需求的学生的成功，包括那些从当地社区大学转到TCNJ的学生。通过与开放科学网格合作访问ELSA，正在刺激研究计划并培养集体影响的联盟。TCNJ ELSA集群是一个异类HPC集群，位于TCNJ HPC中心。它是一种最先进的资源，将继续满足TCNJ理科教师和本科生当前和未来的计算需求。通过该奖项提供的增强将包括购买高端GPU、具有快速互连和大内存容量的快速、现代CPU，以及基于网络的高速存储。升级将直接惠及18名TCNJ教职员工的研究项目，使他们能够继续让本科生参与变革性的研究体验。设计一个同时将GPU节点和CPU节点连接到高速存储的系统的目标是使ELSA能够运行一组不同的研究工作流，以反映TCNJ教员进行的各种跨学科计算研究。这些教师的工作跨越了一系列跨学科的主题，包括(1)计算物理，(2)数学/计算生物学，(3)基因组学，(4)机器学习，以及其他领域。该集群将支持的各种科学努力的一些例子包括细菌菌毛生物力学的分子模拟研究，评估环行双星的宜居性，使用数学模型探索海洋无脊椎动物游泳性能的进化权衡，使用基因组方法表征植物抵御害虫的新调节因子，以及了解如何在机器学习中降低训练数据注释成本。最终，在这个项目中对ELSA集群的增强将显著提高科学发现的能力，并帮助TCNJ教员为本科生做好准备，以便在他们的职业生涯中利用未来日益强大的HPC资源。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

In this project, The College of New Jersey (TCNJ) will acquire equipment to significantly upgrade and enhance our Electronic Laboratory for Science and Analysis (ELSA) High Performance Computing (HPC) cluster. As a primarily undergraduate institution, TCNJ is nationally recognized for the engagement of undergraduate students in research. School of Science faculty work closely with undergraduates in their laboratories throughout the academic year and during the summer in the Mentored Undergraduate Summer Experience (MUSE), with more than 75% of science graduates from TCNJ obtaining at least a semester of research experience. Over the course of this project, nearly 100 undergraduate student researchers per year will benefit by engaging in faculty-mentored research in labs from all School of Science departments (Physics, Chemistry, Biology, Mathematics and Statistics, and Computer Science) as well as in Civil Engineering. These research experiences are transformative for the students enabling their training in computing as well as facilitating their research in fields that include machine learning, astrophysics, biophysics, mathematical biology, and bioinformatics. Beyond the research laboratory, science faculty incorporate ELSA in their teaching, exposing 800-1000 undergraduates per year to advanced computing. TCNJ is also committed to creating a research-intensive environment for all STEM students by increasing success among underrepresented students and those with high financial need, including those that are transferring to TCNJ from local community colleges. The access to ELSA through collaboration with Open Science Grid is stimulating research programs and fostering alliances for collective impact.The TCNJ ELSA cluster is a heterogeneous HPC cluster housed in the TCNJ HPC Center. It is a state-of-the-art resource that will continue to meet the current and future computational needs of TCNJ’s science faculty and undergraduate students. The enhancements provided through this award will include the acquisition of high-end GPUs, fast, modern CPUs with fast interconnects and large memory capacities, and high speed network-based storage. The upgrades will directly benefit the research programs of 18 TCNJ faculty members, allowing them to continue to engage undergraduate students in transformative research experiences. The objective in designing a system with both GPU nodes and CPU nodes connected to high speed storage is to enable ELSA to run a diverse set of research workflows that reflect the varied and interdisciplinary computational research carried out by TCNJ faculty. The work of these faculty spans a range of interdisciplinary themes including (1) computational physics, (2) mathematical/computational biology, (3) genomics, (4) machine learning, and other areas. Some of the examples of the diverse scientific efforts that the cluster will support include molecular simulation studies of bacterial pilus biomechanics, assessing the habitability of circumbinary planets, using mathematical models to explore evolutionary tradeoffs in swimming performance across marine invertebrates, employing genomic approaches to characterize novel regulators of plant defenses against pests, and understanding how to reduce training data annotation costs in machine learning. Ultimately, the enhancements to the ELSA cluster in this project will significantly improve capacity for scientific discovery, and help TCNJ faculty prepare undergraduate students to leverage the increasingly powerful HPC resources of the future in their careers.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.