preview

Analysis Of Parallel And Distributed Computing Via Transforming Parallel Runtimes Into Operating System Kernels

Better Essays

A novel approach to Parallel and Distributed computing via Transforming Parallel Runtimes Into Operating System Kernels Vrishabh A. Lakhani, Deeksha Razdan K.J. Somaiya College of Engineering Information Technology Author Note Abstract General purpose Operating Systems provide services that do not line up with the needs of parallel runtime systems and sophisticated languages and compilers. In an environment that supports the bifurcation of a system into user and kernel level modes, runtime executions at user level cannot leverage hardware features due to requirement of kernel mode privileges. As a result a large portion of the functionality is lost. We thereby propose an idea to transform parallel runtimes into OS kernels. We further hypothesize that it is feasible to do so. To support our hypothesis we also argue that the challenge of future multicore hardware is best met by embracing the networked nature of the machine, rethinking OS architecture using ideas from distributed systems. We investigate a new OS structure, the multikernel, that treats the machine as a network of independent cores, assumes no inter-core sharing at the lowest level, and moves traditional OS functionality to a distributed system of processes that communicate via message-passing. We propose on implementing a multikernel framework model and evaluate such it’s prototype on multicore systems. It may lead to better performance even on present-day machines; comparable with a conventional OS,

Get Access