阿拉巴马大学建全州共享高性能计算中心:Dell与英伟达基础设施成核心

美国阿拉巴马州塔斯卡卢萨县塔斯卡卢萨市消息,Dell Technologies 于 11 月 11 日宣布,阿拉巴马大学(The University of Alabama)新建的高性能计算中心将以 Dell 的高性能计算、超级智能(SI)与数据基础设施为核心,为全州研究者提供大规模计算与数据支持。
这座中心定位为全州共享资源:除阿拉巴马大学系统各校区外,还向阿拉巴马州其他高校、政府机构与产业伙伴开放协作。中心支持的研究领域包括水资源管理、材料与制造、健康、国家安全与量子计算,并为该校新成立的数据科学学院提供计算基础设施。校方此前介绍称,该中心为千万亿次级(petascale)计算系统,目标是让州内研究者能就近使用以往多见于国家实验室与联邦大型计算设施的算力。
在技术构成上,中心采用 Dell PowerEdge 服务器与英伟达(NVIDIA)超级智能基础设施:PowerEdge R770 与 XE7740 系统搭载英伟达 RTX PRO 6000 Blackwell Server Edition 与 H200 NVL 图形处理器,以英伟达 Quantum-2 InfiniBand 高速网络互连,存储由 Dell PowerScale 与 Dell AI Data Platform 文件存储承担,Dell 高性能计算部署服务负责整体实施。Dell 与阿拉巴马大学的合作关系已延续二十多年。
资金方面,该中心由美国商务部国家标准与技术研究院(NIST)部分资助,联邦奖助编号为 60NANB23D149,阿拉巴马大学官方网站对该中心的介绍页标注了同一奖助编号;州政府投入亦是项目资金来源之一。
阿拉巴马大学高性能计算中心执行主任 Fernanda Foertter 表示,这座中心起步于校内对先进计算的需求,州政府的投入让学校得以把目标放大到全州共享,目标是为全州的研究者、学生、政府与产业界提供更易用的先进计算资源。Dell Technologies 高级副总裁 Fran Bogle 表示,这套集高性能计算、超级智能与数据基础设施于一体的平台,将为阿拉巴马州的研究发现、产业协作与长期经济发展打下基础,并能随科学进展持续演进。
(来源:Dell Technologies)
University of Alabama Builds Statewide High Performance Computing Center on Dell and NVIDIA Infrastructure
Tuscaloosa, Tuscaloosa County, Alabama — Dell Technologies announced on November 11 that the University of Alabama's new High Performance Computing Center will run on Dell high performance computing, super intelligence (SI) and data infrastructure at its core, providing large-scale computing and research data support for researchers across the state.
The center is designed as a shared statewide resource. Beyond the campuses of the University of Alabama System, it opens collaboration opportunities to other academic institutions, government agencies and industry partners throughout Alabama. Research fields it supports include water resource management, materials and manufacturing, health, national security and quantum computing, and it will also provide computing infrastructure for the university's new School of Data Science. The university has described the facility as a petascale computing system, intended to give researchers in the state local access to computing capabilities that have historically been available primarily through national laboratories and major federal computing facilities.
Technically, the environment combines Dell PowerEdge servers with NVIDIA AI infrastructure. Dell PowerEdge R770 and XE7740 systems equipped with NVIDIA RTX PRO 6000 Blackwell Server Edition and NVIDIA H200 NVL GPUs are connected by NVIDIA Quantum-2 InfiniBand networking, with Dell PowerScale and the Dell AI Data Platform providing file storage, and Dell HPC Deployment Services supporting implementation of the overall environment. Dell's relationship with the University of Alabama spans more than two decades.
The facility is funded in part by the U.S. Department of Commerce's National Institute of Standards and Technology (NIST), under Federal Award No. 60NANB23D149; the University of Alabama's official page for the center carries the same award number. State of Alabama investment is also among the project's funding sources.
"This center started with a need for advanced computing at the University of Alabama, but the investment from the State of Alabama gave us an opportunity to think much bigger," said Fernanda Foertter, executive director of the University of Alabama High Performance Computing Center. "Our goal is to make advanced computing easier to access and useful to the people solving Alabama's most important research and technology problems." Fran Bogle, senior vice president at Dell Technologies, said the platform, combining HPC, AI and data infrastructure, gives researchers across Alabama the foundation for discovery, industry collaboration and long-term economic development, with the flexibility to evolve with advancements in science.
(Source: Dell Technologies)
来源:Dell Technologies
发布:GSM·美国(发布链接)
图片来源:图:阿拉巴马大学高性能计算与数据中心机房效果图。阿拉巴马大学官方网站(hpc.ua.edu)。 / Photo: Rendering of the data hall at the University of Alabama High Performance Computing and Data Center. University of Alabama official website (hpc.ua.edu).
GSM Global · 本篇英文版 English edition of this story

