ITR/SY: SmartApps: An Application Centric Approach to Scientific Computing

最先进的运行时系统不适合各种动态分布式应用程序，因为它们的设计目的是为各种各样的应用程序提供支持，而没有针对个别特定需求进行大量定制。很少或根本没有信息流从应用程序流向运行时系统，从而使后者能够完全根据应用程序定制其服务。结果，表现令人失望。为了解决这个问题，这个项目将追求以应用程序为中心的计算，或智能应用程序（SMARTAPPS）。SMARTAPPS的首要理念是“衡量、比较，并在有益的情况下进行调整”。也就是说，应用程序将持续监视其性能和可用资源，以确定应用程序是否可以通过重组来改进性能，以及改进多少。然后，如果潜在的性能收益超过了预计的开销成本，应用程序将相应地重构自身和底层系统。这个过程不断发生，以适应应用程序的动态需求和系统资源的可用性。适应可以发生在不同的层次上，包括选择适合当前问题的算法方法、运行时并行化和其他相关的编译器优化、调优可重构操作系统服务（例如调度策略）和系统配置（例如选择使用哪些计算资源）。SMARTAPPS框架提供了性能监控和建模组件，以及执行实际重构的机制，以集成这些级别的适应。该项目将在整个开发过程中在几个重要的计算科学应用程序和代表性平台上测试SMARTAPPS。从计算物理学，它将研究模拟亚原子粒子传输的离散坐标代码。在计算生物学方面，它将研究分子动力学代码和模拟蛋白质折叠和配体结合（也称为药物对接）的代码。SMARTAPPS将在当前的分布式异构平台上开发和运行，包括ASCI机器、大型NUMA机器、工作站网络，以及最终的网格。

State-of-the-art run-time systems are a poor match to diverse, dynamic distributed applications because they are designed to provide support to a wide variety of applications, without much customization to individual specific requirements. Little or no information flows from the application to the run-time system to allow the latter to fully tailor its services to the application. As a result, the performance is disappointing. To address this problem, this project will pursue application-centric computing, or Smart Applications (SMARTAPPS). The overriding philosophy of SMARTAPPS is "measure, compare, and adapt if beneficial." That is, the application will continually monitor its performance and the available resources to determine if, and by how much, the application could improve its performance by restructuring. Then, if the potential performance benefit outweighs the projected overhead costs, the application will restructure itself and the underlying system accordingly. This process occurs continuously to adapt to the dynamic needs of the application and the availability of system resources. The adaptation can occur at various levels including selection of an algorithmic approach suitable for the current problem, run-time parallelization and other related compiler optimizations, tuning reconfigurable OS services (e.g. scheduling policy), and system configuration (e.g., selecting which computational resources to use). The SMARTAPPS framework provides performance monitoring and modeling components, as well as mechanisms for performing the actual restructuring, to integrate these levels of adaptation.This project will test SMARTAPPS on several important computational science applications and representative platforms throughout the development process. From computational physics, it will study discrete-ordinates codes that simulate subatomic particle transport. From computational biology, it will study a molecular dynamics code and a code that simulates protein folding and ligand binding (also known as drug docking). SMARTAPPS will be developed and run on current distributed, heterogeneous platforms, including the ASCI machines, large NUMA machines, networks of workstations, and, ultimately, the Grid.