LIBJACKET v1.0 est maintenant disponible !                                                        

Les fonctions de la bibliothèque LibJacket s'intègrent très facilement dans vos codes C/C++, ceci grâce à son interface matricielle. Elles peuvent également s'intégrer dans vos développements CUDA existants.

Exemple d'utilisation

Estimation de Pi par une méthode de Monte-Carlo :

      #include <stdio.h>
      #include <jacket.h>
      using namespace jkt;
      int main() {
        int n = 20e6; // 20 million random samples
        f32 x = f32::rand(n,1), y = f32::rand(n,1);
        // how many fell inside unit circle?
        float pi = 4.0 * sum_vector(sqrt(x*x + y*y) < 1) / n;
        printf("pi = %g\n", pi);
        return 0;
      }    

• Les clients Jacket existants, bénéficient de 50% de remise sur l'achat d'une licence LibJacket.

• Pour chacun de vos achats d'une carte Nvidia Tesla, vous recevrez une version gratuite de libJacket valable 6 mois.

Autres liens d'informations complémentaires :

• Tour
• Documentation
• Function benchmarks

Through it's matrix interface, you get access to hundreds of functions developed by GPU computing experts at AccelerEyes. Not only is libJacket fast, it's also very comprehensive. It supports all the important data types from double precision to complex numbers. Manipulating vectors, matrices, and N-dimensional arrays is simple with libJacket. Most importantly, all the important routines you would require when working in linear algebra, statistics, imaging, signal processing, computational finance, and other diverse fields are included by default. 

LIBJACKET outperforms equivalent CPU functions. With thousands of lines of optimized code and intelligent algorithms that wisely schedule computations, you will find that the arithmetic intensity of your applications will greatly benefit by using libJacket. Instead of investing time reinventing the wheel with expensive and highly complicated internal development efforts, just grab libJacket!

Benchmarks Using LIBJACKET