CNS Core: Small: Operating System Approaches for a Consolidated Rack Computer

该项目的目标是为一种被称为“机架式计算机”的新型计算机奠定基础。机架式计算机以当今数据中心中通常用来放置计算机的机架命名，它由几台单独的计算机组成，它们共同位于一个机架中。然而，在用户看来，机架式计算机不是大量单独管理的计算机，而是具有单个计算机的组合资源的单个计算机。为了实现这一目标，PI将研究新的编程模型、通信机制和操作系统设计。与云计算、集群计算和超级计算领域的现有技术相比，机架式计算机允许相同的程序更无缝地从基本笔记本电脑扩展到功能强大的服务器的完整机架。由于一个机架可以容纳大量的计算能力，包括多达10,000个CPU核心，而且机架计算机与当前技术相比提供了显著的效率提升，因此一台机架计算机可能足以承载许多服务，这些服务目前需要复杂的工程才能实现互联网规模。这承诺大幅降低开发和部署的复杂性，从而降低成本，同时提高能源效率。这一奖项反映了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 objective of this project is to lay the foundations for a new type of computer, termed a “rack computer”. Named for the racks commonly used to hold computers in today's data centers, a rack computer consists of several individual computers, co-located in a rack. Instead of large number of individually managed computers, however, a rack computer appears to the user as if it was a single computer with the combined resources of the individual machines. Toward this goal, the PI will investigate new programming models, communication mechanisms and operating system designs. Compared to existing techniques from the cloud-, cluster- and super-computing worlds, a rack computer allows the same program to more seamlessly scale from a basic laptop computer, to a full rack of powerful servers. Because one rack can hold a substantial amount of computing power, including as many as 10,000 CPU cores, and because rack computers offer substantial efficiency gains over current techniques, a single rack computer may be enough to host many services that today require complex engineering to bring to Internet scale. This promises to substantially reduce both development and deployment complexity, and thus costs, while at the same time improving energy efficiency.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.