Intel® Parallel Studio XE

yet another ifort 18 regression with OpenMP and allocatables (invalid read in scalable_free, frontend.cpp:2868)

Dear Intel team,

it seems something is fundamentally broken when using allocatables in connection with OpenMP in ifort 18.x.

On top of my earlier bug reports #03202444 and #03214078, here is another related problem:


Math intrinsic performance issue with ifort17

After migrating my code from ifort15 to ifort17 I observe a substantial slowdown caused by the pow() intrinsic. The runtime profile shows __bwr_pow with almost 3x the sample count compared to the previous pow.L. Also the SVML version is notably slower. The other routines have not changed much.

Build and profiling were done on a Haswell running RH6. Relevant FP compile switches are:

-qopenmp -O3 -fp-model fast=1 -r8 -xAVX -fp-model fast=2 -no-prec-div -no-prec_sqrt -ftz -fast-transcendentals

Profile with ifort15 exe:

Application crashing on computers with certain CPU (i7-8700K)

We have a customer who bought Alienware desktops with i7-8700K processors and reported that our (console) application is randomly crashing on them - the console window disappears. Alienware provides a tool called Support Assist to check system event and our application's crashes are shown there with some details, e.g. fault bucket number and type, WER (Windows Error Reporting) data.

My main question is whether there are any compiler or linker settings in IVF that might cause application crashes on computers with certain CPUs.

Visual Studio Version for older Visual Fortran

I have the following Fortran package:

Intel® Parallel Studio XE Composer Edition for Fortran Windows* (incl. Microsoft Visual Studio Shell and Libraries*) - 2013 SP1 Update 6.

I have been using it with the built in shell.  I would like to get a real version of Visual Studio and integrate the Fortran compiler into that.  I gather that it is not compatible with the 2017 version.  Can you tell me what the newest version is that it will work with?




VS 2017 Preview

I switched over to VS 2017 Preview for some C# features. Now I find that compiling my Intel Fortran programs I run into an error that it cannot find the ucrtd.lib for Windows.  I can find the lib and link it directly without trouble but I am just to quick to get the latest VS.


Update 6 will not complete the install

I've got Intel Parallel Studio XE 2017 and ISM says Update 6 is available. I got our LAN police to grant me rights to my PC so I can install. I've hit the Update button 4 times today and the package seems to download but the Install never finishes. The blue box pops up and some XML stull seems to be doing something, and at one point it asks if I want to make changes and I say yes, then the ISM box is still there so I figure I'll hit the Update button again. The download goes again and the process repeats.

Possibility to parallelize Cholesky decomposition routine?

I have old FORTRAN code for Cholesky decomposition of symmetrical FEM matrix (attached file). Subroutine seems to be the slowest part of conjugate gradient solver. I want to parallelize it, but it seems to me that it is impossible.  Number of unknowns (variable "is") is typically 200 thousands to several millions. Variable "i2" is typically bellow 1000.

Has anybody have an idea how to rewrite routine to enable its parallel processing?

Thanks for any advice

OMP: Info #261

Dear all,
When link with latest IVF 2018, the following information displayed when running the program:

OMP: Info #261: KMP_ALL_THREADS variable deprecated, please use KMP_DEVICE_THREAD_LIMIT instead

How to set up environment to remove this information?

Tang Laoya

Difficulty in tracking down the source of a NaN

I have some code which sometimes produces NaN results for some elements of an array. Interestingly, this happens when I enable -O2, but not when I enable -O3. It doesn't seem to happen with gfortran under linux either. Of course, there may be some underlying issue with my algorithm which makes it susceptible to changes in compiler or optimisation level.

Subscribe to Intel® Parallel Studio XE