Special Issue on Emerging Many-Core Systems for Exascale Computing
Abstract
Special Issue on Emerging Many-Core Systems for Exascale Computing 1. INTRODUCTION To scale computing systems to Exaflops, future many-core systems will require the integration of hundreds of cores on a single chip. The large number of cores in emerging many-core systems introduces complexities associated with resource sharing, parallel programming, heterogeneity, distributed memory architectures, and implementation challenges such as 3D stacking. The design of such architectures will require innovations in all areas of circuits and systems. This special issue aims to present and discuss innovative ideas and solutions in the design, modeling, prototyping, programming, and implementation of many-core systems for exascale computing. In what follows, we will give a brief overview of the accepted articles in this special issue based on their topics. 2. SELECTED ARTICLES In the era of reconfigurable platforms for exascale computing hosting multiple applications with arbitrary performance requirements, providing worst-case platform-wide voltage/frequency operating point is neither optimal nor desirable. As a solution to this problem, designs commonly employ dynamic voltage and frequency scaling (DVFS), promising significant energy and power reductions by providing each application with the operating point tailored to its needs. To further enhance the optimization potential, recent works interleave dynamic parallelism with conventional