ABSTRACT
This chapter describes the challenges introduced by the need for time-predictability on the one hand, and high-performance on the other, and discusses a high level how these contradictory requirements can be simultaneously supported. High-performance computing has been for a long time the realm of a specific community within academia and specialized industries; in particular those targets demand analytics and simulations applications that require processing massive amounts of data. Until now, trends in high-performance and embedded computing domains have been running in opposite directions. In order to enable predictable parallel performance to be analyzed, it is required that the application parallel graph is known, with control- and data-flow information needed for the analysis of the timing behavior of the parallel program. The use of multi- and many-core platforms considerably challenges approaches for real-time timing analysis, required to determine worst-case execution time of the application code. The chapter also presents an overview of the key concepts discussed in this book.
