• State Accepted Answer
  • +1 person also asked this people also asked this
  • Locked Locked
  • Replies 2 replies
  • Subscribers 19 subscribers
  • Views 10370 views
  • Users 0 members are here
Options
Related
How was your experience today?
This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

Optimized ARM  linear algebra library

Mazen Ezzeddine

Dear all,

- Does ARM provide any highly optimized linear algebra library  such as Intel MKL? can I use it for academic purposes?

- I need to compile a library and make sure that it uses wherever possible SIMD vector instructions, what exact command line nee I to provide to the gcc compiler for such a purpose? any hint?

kind thanks

  • 0

    Hi Mazen,

       You could try the Eigen Library.

    It uses expression templates to generate very optimized NEON instructions from high level algebra in C++ and it definitely compiles with gcc 4.8 and up.

    Cheers

    Ben

  • +1

    Thanks much bensch128,

    I will surely look at Eigen Library and perhaps link my code against it; however, do you have an idea of what advantages it provides as compared to ARM-based openBLAS and  ATLAS that I am currently linking against?

    Also, I am linking  to the  linaro/ubuntu distribution of openBLAS and ATLAS  which are compiled with hardware floating point support, are they already compiled for usage of neon vectorized instructions or just typical hardware floating point instructions? any idea/hint?

    Best wishes.