The dynamics of disk clearing, late stage planetesimal disk and planetary system evolution

该奖项由 2009 年美国复苏和再投资法案（公法 111-5）资助。 Alice Quillen（罗切斯特大学）将领导行星形成早期阶段轨道动力学的理论研究。 她将重点研究原行星和小天体的引力相互作用，并将尘埃的产生和形态与行星生长、迁移、散射和共振交叉等过程联系起来，并探索如何观察这些过程。 与之前的研究相比，这项研究将增加模拟粒子的数量和质量范围，并将考虑碰撞演化、吸积和引力相互作用。这将需要增强能力的计算机代码，因此该小组将开发在图形处理单元上执行并行计算的方法。这些单元有几个潜在的优势，主要是它们是低成本平台，结合了许多处理器和快速且大的设备内存。 Quillen 博士和她的团队将为暑期研究生和教师开发并举办一个关于显卡并行计算的年度研讨会。学生将学习使用 CUDA（统一计算设备架构）编程语言编写、测试和执行简单的程序。 与往年一样，新的研讨会将接待来自多个部门的学生和研究人员；这将为跨学科合作提供机会，促进跨学科思想和算法的转移，并最大限度地提高学生的参与度和研究效率。还将为编程经验有限的学生举办其他介绍性研讨会。

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5).Dr. Alice Quillen (University of Rochester) will lead a theoretical study of the orbital dynamics of the early phases of planet formation. She will focus on the gravitational interactions of protoplanets and small bodies, and will relate dust production and morphology to processes such as planet growth, migration, scattering, and resonance crossing, and explore how these processes could be observed. This study will increase the numbers and mass range of particles simulated, compared to previous efforts, and will consider collisional evolution, accretion, and gravitational interactions. This will require computer codes of increased capability, so this group will develop methods for performing parallel computations on graphics processing units. These units have several potential advantages, mainly that they are low-cost platforms combining many processors with fast and large on-device memory. Dr. Quillen and her group will develop and run a yearly workshop on parallel computing with graphics cards for summer research students and faculty. Students will learn to write, test, and execute simple programs in the CUDA (Compute Unified Device Architecture) programming language. As in previous years, new workshops will host students and research faculty from multiple departments; this will provide opportunities for interdisciplinary collaboration, promote transference of ideas and algorithms across disciplines, and maximize student involvement and research productivity. There will also be other workshops of an introductory nature for students with limited programming experience.