Collaborative Research: CSR-PSCE, SM: Adaptive Memory Management in Shared Environments

程序性能在很大程度上取决于程序可用的内存量。在传统的计算系统中，应用程序的内存工作集有一个有限的大小--为应用程序提供更多的内存可以提高性能，直到它的工作集得到满足。一旦满足工作集，增加的内存几乎不会带来任何好处。然而，在垃圾收集(一种内存管理技术，其中不太可能被应用程序重用的空间被自动回收)的存在下，程序性能和内存分配之间的关系更加复杂。数据在三个级别进行管理：编译器在程序级别管理数据对象，垃圾收集器在虚拟机级管理堆，虚拟内存管理器在操作系统级别管理虚拟内存。中间层起着关键作用。增加应用程序的堆大小可以减少垃圾回收的频率并提高性能，但太大的堆可能会触发分页并降低性能。

Program performance is highly dependent on the amount of memory available to theprogram. In traditional computing systems, the memory working set of an applicationhas a bounded size - providing more memory to an application improves performanceuntil its working set is met. Once the working set is met, additional memory yields littleor no benefit. However, in the presence of garbage collection (a technique for memorymanagement where space that is unlikely to be reused by an application isautomatically reclaimed), the relationship between program performance and memoryallocation is more complex. Data is managed at three levels: the compiler managesdata objects at the program level, the garbage collector manages the heap at the virtualmachine level, and the virtual memory manager manages virtual memory at theoperating system level. The middle layer plays a critical role. Increasing an application'sheap size can reduce the frequency of garbage collections and improve performance,but too large a heap may trigger paging and degrade performance.