We present a new source-to-source transformation which generates code to compute several model instances simultaneously. Due to the increased memory locality of memory accesses this speeds up the computation on processors using a cache hierarchy to overcome the relative slow memory access. The speedup depends on the model code, the processor, the compiler, and on the number of instances. © 2012 Springer-Verlag.
CITATION STYLE
Giering, R., & Voßbeck, M. (2012). Increasing memory locality by executing several model instances simultaneously. In Lecture Notes in Computational Science and Engineering (Vol. 87 LNCSE, pp. 93–101). https://doi.org/10.1007/978-3-642-30023-3_9
Mendeley helps you to discover research relevant for your work.