Intel® Fortran Compiler
Build applications that can scale for the future with optimized code designed for Intel® Xeon® and compatible processors.
Announcements
The Intel sign-in experience is changing in February to support enhanced security controls. If you sign in, click here for more information.

LAPACK linking

Ashkan_K_
Beginner
772 Views

Dear all,

I'm trying to use LAPACK with Intel Visual Fortran Compiler for Windows. I'm hoping to be bale to use it in ABAQUS  UEL subroutine.

Here is the guideline to install LAPACK (http://icl.cs.utk.edu/lapack-for-windows/lapack/index.html#libraries):

  1. Download the LAPACK precompiled binaries. File names of the precomputed debug libraries end with the letter "d" e.g. BLASd.lib and LAPACKd.lib (in comparison to the release versions BLAS.lib and LAPACK.lib).
  2. Locate your BLAS libraries for your machine. (You may want to choose the Debug config if you choose GOTOBLAS)
  3. Download the LAPACK-VS-Example Visual Studio project and unzip
  4. Move or Copy the libraries from step 1 in the LAPACK-VS-Example folder.
  5. If you are not using the Reference BLAS, you will need to modify change the properties "Linker > General > Additional Library Directory" to tell Visual Studio where the libraries are, and also add the name of your BLAS library in "Linker > Input > Additional Dependencies"
  6. Compile the project and run the resulting executable.

Because I'm using the reference BLAS, I'm not suppose to deal with step 5.

I have two questions:

1- Is it possible to use LAPACK with ABAQUS UEL subroutine or not?

2- How can I locate BLAS libraries to my machine (step 2)? without locating them to machine (which I don't know how to do) I'm getting these errors while running it:

Error    1     error #11018: Cannot open LAPACK.lib    ipo 

Error    2     error #11018: Cannot open BLAS.lib    ipo    

Error    3     fatal error LNK1181: cannot open input file 'LAPACK.lib'    LINK    

Thanks in advance for helping me.

Regards,

   

 

0 Kudos
8 Replies
TimP
Black Belt
772 Views

You might start by using the MKL library which is included with ifort (optional install step).  You should have noticed that the utk instructions suggest this.  The /Qmkl options provide an easy way to deal with linking.  Given that ifort supports this method, you shouldn't be surprised if the utk libraries don't get much support on this forum.  As Abaqus versions set up for ifort UEL would also use MKL, you could bring grief upon yourself by choosing a conflicting library implementation.

Ashkan_K_
Beginner
772 Views

Dear Tim,

I really appreciate your consideration.

Should I have buy this MKL? If the asnwer is yes, I'm wondering if there's any other way to do that or not?

I'm under pressure because of an upcoming deadline and It would be great if you can help me a bit more. I should mention that I'm a new user of Fortran and LAPACK.

I'm looking forward to hearing from you.

Best Regards,

Steven_L_Intel1
Employee
772 Views

You already have MKL - it's part of Intel Visual Fortran. You can ask in the MKL forum if you need more help.
 

mecej4
Black Belt
772 Views

A point of clarification and emphasis: MKL is a replacement for both BLAS and Lapack, whereas Netlib and Netlib-origin libraries may keep the two separated (for historical reasons -- BLAS was created a few years before Lapack).

Ashkan_K_
Beginner
772 Views

Dear Mecej4,

So you mean that instead of using LAPACK, I can use MKL. Am I right?

Does MKL have functions for eigenvalue, eigenvector, SVD, ...?

Best Regards,

mecej4
Black Belt
772 Views

Ashkan K. wrote:

So you mean that instead of using LAPACK, I can use MKL. Am I right?

Yes.

Does MKL have functions for eigenvalue, eigenvector, SVD, ...?

Yes. In addition, MKL includes routines for problem types that are not part of the scope of Lapack, such as FFT, nonlinear least squares, etc.

Steven_L_Intel1
Employee
772 Views

It might be easier to understand that MKL includes the LAPACK functions, in implementations that are parallelized and optimized.

Ashkan_K_
Beginner
772 Views

Tanks everybody to be such a helpful.

I really appreciate it.

Reply