The power consumption of programs and algorithms is currently a very active research field. This includes the investigation effect different programming techniques on consumption. Some have already been studied intensively. However, there are that did not get as much attention needed so far. One these vectorization programs, which uses special operations to calculate several data in one step. In this article, we investigate study program versions dense matrix multiplication combine with...

The combination of different parallel programming environments can be used to exploit all heterogeneous levels hardware, which might lead an optimization application programs. An exemplary is the use Message Passing Interface (MPI) together with vectorization based on Advanced vector extensions (AVX), investigated in this article. A special emphasis lies MPI data orderings and their influence AVX strategies. Jacobi method as case study for several program version have been implemented analyzed.

The energy efficiency of program executions is an active research field in recent years and the influence different programming styles on consumption part effort.In this article, we concentrate SIMD study effect vectorization performance as well power consumption.Especially, programs using AVX instructions are considered focus load store instruction set.Several semantically similar but selected used to build versions for same algorithm.As example application, Gaussian elimination has been...

Modern computer environments are limited by energy and power constraints during the execution of programs. These limits can be due to lines, budgeting, ecology, battery life or many other reasons. To bypass these limits, hardware software development strive reduce consumption algorithms. This article investigates capabilities limitations vectorization with respect efficiency. Vectorization is a technique instrument on-chip SIMD that increases performance The capability programs has shown. As...

Abstract This article provides an enhanced parallelization of the WSM7 microphysics scheme for Weather Research and Forecasting Model (WRF). The is designed to maximize utilization a heterogeneous computing system consisting CPUs, GPUs or both. Therefore reference implementation re-implemented execution model. For each time step, dynamic load distribution introduced which balances computational between two components aiming overall minimum time. evaluation parallelized done specific weather...

In the new era of cyber-physical systems, software must adapt itself to ever-changing environmental conditions and situations. This is currently not reflected in design embedded operating since they are primarily optimized for fixed usage scenarios with tight resource constraints. We discuss idea interpreted system kernels, which can form a foundation highly reconfigurable systems. The paper elaborates reasonable use cases, shows comparable approaches from past sketches an implementation...