MRI: Acquisition of Artificial Intelligence Super Computer (AISC) for Accelerating Scientific Discovery

该项目出资购买了一台人工智能超级计算机(AISC)，这将使研究人员、企业家、教育工作者和政策制定者能够利用以前无法想象的资源，根据数据做出明智的决策。AISC将推动依赖人工智能和机器学习(AI/ML)技术以及高性能计算(HPC)技术的众多科学、工程和医疗领域的变革性进展。AISC将支持来自凯斯西储大学及其合作机构的研究人员。AISC支持的研究范围可分为四大主题：(I)网络基础设施(CI)和计算机科学；(Ii)材料科学；(Iii)工程系统；(Iv)生物医学工程。除了支持广泛的研究，该项目还将围绕加速AI/ML计算建立一个充满活力的多学科社区，并将举办课程、教程和研讨会，在当地和全国范围内吸引计算机科学和学科科学的本科生、毕业生、博士后和教职员工。AISC采用具有大容量、快速和高带宽内存的张量核心图形处理单元(GPU)，以及基于非易失性内存Express(NVMe)的存储和高速互联等尖端技术。AISC可以划分为多个实例，每个实例都具有各自的高带宽内存、高速缓存和计算核心，允许支持不同大小的作业，并保证每个作业的服务质量(Qos)，从而优化利用率，从而将AISC的覆盖范围扩展到更多用户。AISC的更广泛影响不仅包括研究成果和新技术，还包括将成果转化为实践，并将进展广泛传播到CI、AI和特定科学领域以外的广泛社区。为了充分实现该项目更广泛的影响愿景，设想了以下活动：(I)培养STEM人才并提高公众对科学和计算的兴趣；(Ii)在促进教学、培训和学习的同时促进发现和理解；(Iii)熟悉人工智能的CI劳动力发展；(Iv)每年特定领域的黑客松；以及(V)通过与NSF赞助的多个主要项目合作，扩大代表不足的群体的参与。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。

The project funds the acquisition of an Artificial Intelligence Super Computer (AISC) which will enable researchers, entrepreneurs, educators, and policymakers to leverage previously unimaginable resources to make intelligence-¬informed decisions from data. AISC will enable transformative advances in myriad scientific, engineering, and healthcare fields that rely on artificial intelligence and machine learning (AI/ML) techniques and high-performance computing (HPC) technologies. AISC will support researchers from across Case Western Reserve University and their collaborating institutions. The spectrum of research enabled by AISC can be classified into four broad themes: (i) Cyberinfrastructure (CI) and computer science; (ii) material science; (iii) engineering systems; and (iv) biomedical engineering. In addition to supporting a broad spectrum of research, the project will build a vibrant multidisciplinary community around accelerating AI/ML computing and will conduct classes, tutorials, and workshops to engage undergraduates, graduates, post-docs and faculty in computer science and discipline sciences within the local Cleveland area and nationally.AISC adopts Tensor Core graphics processing units (GPUs) with large, fast, and high bandwidth memory along with cutting edge technologies such as Non-Volatile Memory Express (NVMe) based storage, and a high-speed interconnect. AISC can be partitioned into multiple instances each fully isolated with their own high-bandwidth memory, cache, and compute cores allowing support of varying size jobs with guaranteed quality of service (QoS) for every job, optimizing utilization and thereby extending the reach of AISC to more users. Broader impacts from AISC include not only research findings and new technologies, but transfer to practice and wide dissemination of advances to broad communities beyond the CI, AI and specific scientific domains. To fully realize the project’s broader impact vision, the following activities are envisioned: (i) Build STEM talent and improve public interest in science and computing; (ii) Advance discovery and understanding while promoting teaching, training, and learning; (iii) CI workforce development that is conversant with AI; (iv) Yearly domain-specific hackathons; and (v) Broaden participation of underrepresented groups by partnering with multiple major NSF-sponsored programs.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.