CUBLAS

AcronymDefinition
CUBLASCompute Unified Basic Linear Algebra Subprograms (NVidia)
References in periodicals archive ?
En este sentido, gran parte del esfuerzo inicial para acelerar nucleos de ALN con GPU, hasta la presentacion en 2007 de CUBLAS por parte de NVIDIA, consistio en la implementacion de operaciones incluidas en la especificacion BLAS.
Posterior a la presentacion de CUBLAS destaca el trabajo de Ries et al.
The FFT/IFFT modules have approximately the same performance because similar to [7], both are implemented by the CUBLAS function library.
These products can be performed on a GPU with a function cublasDgemm from the CUBLAS library [28].
Let us recall, that the baseline version implements the algorithm shown in listing 1 using CUBLAS procedures for matrix-matrix products.