Title :
Achieving Natural Mathematical Expression Programming on GPUs via Expression Templates
Author :
Breglia, Alfonso ; Capozzoli, Amedeo ; Curcio, Claudio ; Liseno, Angelo
Author_Institution :
Dipt. di Ing. Elettr. e delle Tecnol. dell´Inf., Univ. di Napoli Federico II, Naples, Italy
Abstract :
We present the development of one of the first libraries based on the so-called expression templates technique to simplify the implementation of CPU and parallel GPU codes. Expression templates allow to express matrix algebra operations to be executed either on the CPU or on the GPU with a syntax very close to the natural mathematical one. The developed library has been deeply optimized so that the same performance of standard C/C++ or CUDA codings is kept.
Keywords :
graphics processing units; mathematics computing; matrix algebra; parallel architectures; software libraries; C/C++; CPU codes; CUDA codings; expression templates; libraries; matrix algebra operations; natural mathematical expression programming; parallel GPU codes; Central Processing Unit; Graphics processing units; Kernel; Libraries; Performance evaluation; Syntactics; Vectors; CUDA; GPU; Metaprogramming; expression templates;
Conference_Titel :
Modelling Symposium (EMS), 2013 European
Conference_Location :
Manchester
Print_ISBN :
978-1-4799-2577-3
DOI :
10.1109/EMS.2013.84
