CSSI: Frameworks: X-Ion Collisions with a Statistically and Computationally Advanced Program Envelope (X-SCAPE)

CSSI：框架：具有统计和计算高级程序包络的 X-Ion 碰撞 (X-SCAPE)

• 批准号: 2004571
• 负责人: Abhijit Majumder
• 金额: $ 404.88万
• 依托单位: Wayne State University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2020
• 资助国家: 美国
• 起止时间: 2020-07-01 至 2024-06-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2004571&HistoricalAwards=false
• 关键词: CSSI; Frameworks; Ion; Collisions; Statistically

High-energy colliders around the world currently study the production of, and interaction between, a variety of subatomic particles and environments produced in the collision of protons with protons, with nuclei, or even between two nuclei, as at the Large Hadron Collider (LHC) at the European Center for Nuclear Research (CERN), or at the Relativistic Heavy-Ion Collider (RHIC) at Brookhaven National Laboratory (BNL). The future Electron Ion Collider (EIC), slated for construction at BNL, will study collisions of electrons with protons and nuclei. Interactions at all these experiments are dominated by the strong nuclear force, whose behavior is still not well understood. Direct comparison between theory and experiment requires sophisticated computer simulations, where each stage of a collision is modeled via a combination of established principles or candidate theories. A complete simulation, consisting of a number of sub-simulations, depends on several input parameters. The extracted values for these parameters represent fundamental properties of strongly interacting matter. A simultaneous determination of these parameters, in extensive comparisons with volumes of data from diverse experiments, requires an elaborate statistical and computational framework. The X-SCAPE collaboration, a multi-disciplinary team of physicists, computer scientists and statisticians, is engaged in the construction of such an open-source framework. The expertise required to operate, modify and extend this framework is disseminated to practicing scientists via a combination of dedicated short-term schools (in-person and virtual), topical workshops, web tutorials, conference presentations and detailed publications.The X-SCAPE project encompasses the entire high-energy nuclear physics enterprise by providing a general purpose, modifiable, and modular framework, that incorporates all known interaction processes prevalent at these experiments. The developed software is designed to be portable onto hybrid architectures, where GPU acceleration can be applied to the most demanding computational tasks. Provided as a steadily improving package culminating in annual releases, X-SCAPE includes of i) the default distribution of baseline tools allowing for simplified simulations on commodity CPUs, ii) a refactored product for distributed architectures with GPUs, and iii) Bayesian statistical routines to both emulate and systematically explore the model space of parameters against diverse experimental data.This award by the Office of Advanced Cyberinfrastructure is jointly supported by the Physics at the Information Frontier (PIF) Program within the Division of Physics and the Office of Multidisciplinary Activities within the Directorate for Mathematical and Physical Sciences.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.

世界各地的高能对撞机目前研究各种亚原子粒子的产生和相互作用，以及质子与质子、原子核甚至两个原子核之间碰撞产生的环境，如欧洲核子研究中心的大型强子对撞机（LHC）或布鲁克海文国家实验室的相对论重离子对撞机（RHIC）。未来的电子离子对撞机（EIC），预定在BNL建造，将研究电子与质子和原子核的碰撞。在所有这些实验中，相互作用都是由强核力主导的，但人们对强核力的行为仍然没有很好的理解。理论和实验之间的直接比较需要复杂的计算机模拟，其中碰撞的每个阶段都通过已建立的原理或候选理论的组合来建模。一个完整的模拟，包括一些子模拟，取决于几个输入参数。这些参数的提取值代表强相互作用物质的基本性质。这些参数的同时测定，在广泛的比较与大量的数据从不同的实验，需要一个精心设计的统计和计算框架。X-SCAPE协作是一个由物理学家、计算机科学家和统计学家组成的多学科团队，致力于构建这样一个开源框架。操作、修改和扩展这一框架所需的专门知识通过专门的短期学校向实践科学家传播（面对面和虚拟），专题研讨会，网络教程，会议演示和详细的出版物。X-SCAPE项目通过提供通用的，可修改的和模块化的框架，它包含了所有已知的在这些实验中普遍存在的相互作用过程。开发的软件被设计为可移植到混合架构上，其中GPU加速可以应用于最苛刻的计算任务。X-SCAPE作为一个稳步改进的软件包，最终在年度发布中达到顶峰，包括i）允许在商品CPU上进行简化模拟的基线工具的默认分布，ii）用于带有GPU的分布式架构的重构产品，以及iii）贝叶斯统计例程，以模拟和系统地探索模型空间的参数对不同的实验数据。这个奖项由高级网络基础设施办公室，该奖项由物理学部内的信息前沿物理学（PIF）计划和数学与物理科学理事会内的多学科活动办公室共同支持。该奖项反映了NSF的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

项目成果

Bayesian Inference of the Fluctuating Proton Shape in DIS and Hadronic Collisions

DIS 和强子碰撞中波动质子形状的贝叶斯推断

• DOI: 10.5506/aphyspolbsupp.16.1-a33
• 发表时间: 2023
• 期刊: Acta Physica Polonica B Proceedings Supplement
• 作者: Mäntysaari, H.;Schenke, B.;Shen, C.;Zhao, W.
• 通讯作者: Zhao, W.

Gaussian Process Subspace Prediction for Model Reduction

用于模型简化的高斯过程子空间预测

• DOI: 10.1137/21m1432739
• 发表时间: 2022
• 期刊: SIAM Journal on Scientific Computing
• 影响因子: 3.1
• 作者: Zhang, Ruda;Mak, Simon;Dunson, David
• 通讯作者: Dunson, David

Quenching of jets tagged with W bosons in high-energy nuclear collisions

• DOI: 10.1103/physrevc.105.054902
• 发表时间: 2021-03
• 期刊: Physical Review C
• 影响因子: 3.1
• 作者: Shan-Liang Zhang;Xin-Nian Wang;Ben-Wei Zhang

First Results from Hybrid Hadronization in Small and Large Systems

• DOI: 10.22323/1.387.0158
• 发表时间: 2020-09
• 期刊: Proceedings of 10th International Conference on Hard and Electromagnetic Probes of High-Energy Nuclear Collisions — PoS(HardProbes2020)
• 作者: M. Kordell;Angerami;S. Bass;S. Cao;Y. Chen;J. Coleman;L. Cunqueiro;T. Dai;L. Du;R. Ehlers;H. Elfner;D. Everett;W. Fan;R. Fries;C. Gale;F. Garza;Y. He;M. Heffernan;U. Heinz;B. Jacak;P. Jacobs;S. Jeon;W. Ke;E. Khalaj;B. Kim;A. Kumar;D. Liyanage;T. Luo;M. Luzum;A. Majumder;M. McNelis;J. Mulligan;C. Nattrass;D. Oliinychenko;L. Pang;C. Park;J. Paquet;J. Putschke;G. Roland;B. Schenke;L. Schwiebert;C. Shen;A. Silva;C. Sirimanna;R. Soltz;Y. Tachibana;G. Vujanovic;X. Wang;R. Wolpert;Y. Xu

Jet transport coefficient q^ in lattice QCD

• DOI: 10.1103/physrevd.106.034505
• 发表时间: 2022-08
• 期刊: Physical Review D
• 影响因子: 5
• 作者: Amit Kumar;A. Majumder;J. Weber