Project News

Implementing S(o)OS

The S(o)OS project primarily focused on generating a suitable architecture for future operating system, but did not have the resources to implement a working kernel. Instead, it generated the necessary set of algorithms, interfaces and protocol descriptions, preparing an implementation as much as possible and allowing evaluation of behaviour and performance.

Developing such a kernel is hard work, in particular if the full functional scope of S(o)OS is to be maintained, which involves multiple different instances of all modules to reflect different protocols, platforms and environments. The MyThOS project, funded by the Federal Ministry for Education and Research (BMBF), is a first attempt towards realising such an implementation, by focusing on a limited subset of capabilities and platforms. Specifically, the MyThOS project will realise a modular base kernel that will allow booting and executing a single large scale application. The main focus thereby rests on improving in particular the performance of individual threads, in terms of instantiation time, execution time, communication overhead etc. The goal will be to provide an implementation of the operating system for dedicated distributed infrastructures that host a highly dynamic, large scale application.

Reference applications that will be supported by MyThOS are molecular simulations, distributed stream processing, electric current simulations. Project partners include: University of Ulm; HLRS, University of Stuttgart; Brandenburg University of Technology, Cottbus; Alcatel-Lucent Germany AG, Bell-Labs; University of Siegen. The MyThOS project will start October 2013 and run for 3 years.