TIMEF() exhibits different behaviour in Linux and Windows
I have spent some time today trying to sort out a bottleneck in my code and been using timef() from the portability library.I have both linux (Ubuntu 16) and Windows (7) versions of the compiler - both...
View ArticleCompiler flags for AMD Epyc processors
I hope that this is a simple question:Which compiler flags should i use to get the best performance out of an AMD Epyc processor (particularly for MPI and OpenMP codes)? I know which instruction sets...
View ArticleHDF5 1.10.1 compiling error with ifort 18.0.0
Sorry to bother you. I have been trying to compile HDF5 version 1.10.1 (downloaded from https://support.hdfgroup.org/HDF5/release/obtainsrc.html) with (the flags --enable-fortran --enable-shared...
View ArticleParallel Studio XE fails to install on Ubuntu (WSL)
Environment InformationWhat platform are you using? (please provide specific distribution/version in summary) Ubuntu 16.04.3 LTS on Windows Subsystem for Linux (WSL) - Windows 10 Version 10.0.16299...
View Articlenew namelist bug for whitespace in derived types in 2017 and 2018
Hi, There is a minor new bug In 2017 and 2018 compilers related to parsing white space in derived types in namelist input. It did not happen in the 2015 version. Here is a simple test case:program...
View ArticleOpenMP and allocatable derived types - still a bad idea?
I need to further increase performance on a grid generation algorithm I wrote. This code is based on an octree data structure, implemented via fortran derived types. So lots of allocatable derived...
View ArticleDifference between COMMITQQ, PXFFFLUSH, and FSYNC
In rare cases when my code crashes soon after formatted IO, some data are not actually written to disk. I am currently calling `FLUSH` on all files (before the crash) but it is apparently insufficient....
View ArticleRefuses to thread OpenMP loop
Hi,I have two OpenMP loops that ifort (Version 16.0.1.150 Build 20151021) refuses to parallelize. I've added "-qopenmp -opt-report-file=stdout -opt-report-phase:openmp " to my compile/link commands...
View ArticleNo MIC MPI installation for 2018 release?
I just removed Intel Parallel Studio Cluster Edition 2017 and installed the 2018 version on a centOS box. However, when I tried to compile Fortran code for MIC, mpifort said it couldn't find...
View ArticleCompile time bug in Ifort 18.
Using ifort version: 18.0.0.20170811Test program:module ifort18_bug implicit none TYPE A type(B),pointer :: w1=> null(),w2 => null() END TYPE A type B integer :: i end typeend...
View Article/ifort-18_0: command line: warning #10006: ignoring unknown option...
I have installed Xcode 8.3.3 and Intel Parallel Studio XE Composer Edition for Fortran macOS 2018. Then, I created an Xcode project - selected the Intel Compiler - added a new Fortran source file....
View ArticleMPI Layout Output Difference: Which Flag(s)?
All,I am exploring the use of Intel Fortran's "go fast" flags to try and make my program "go fast". First, let me say that by default, we build our model with:-O3 -qopt-report0 -ftz -align all...
View ArticleOld -FR option, what is this?
I have a Makefile with ifort option -FR - I can't find this in the current documentation, what is this option? I remember seeing it years ago, perhaps it's been replaced with a new option?
View ArticleCompiler flags for AMD Epyc processors
I hope that this is a simple question:Which compiler flags should i use to get the best performance out of an AMD Epyc processor (particularly for MPI and OpenMP codes)? I know which instruction sets...
View ArticleDerived data type memory access
Hello everyone!We have found strange behaviour of the following simple program that implements jacobi-like iteration with array of derived data type (block_t). This type consists of one-dimensional...
View ArticleIncorrect message when fpe0 is turned on
Sample code:program main IMPLICIT NONE write(*,*) 1D-300*1D-300 write(*,*) 0D0/0D0 end program mainCompile it withifort -debug -traceback -check all -fpe0 filename.f90I get the following output...
View Articleifort 17.0.4.196 hangs at 100% CPU if input from stdin is not newline-terminated
program test implicit none character*200 line 10 read(*,'(a)', end=20,err=30) line write(0,'(a)') line goto 10 20 write(0,'(a)') 'EOF' stop 30 write(0,'(a)') 'ERROR' stop end program testCompiling the...
View ArticleWhat is the default value for FORT_BUFFERING_THRESHOLD?
https://software.intel.com/en-us/node/691978 only says that the value should be "an integer value greater than zero and less than 2147483647". But, there is no mention of the default value.
View ArticleFortran Precision
I've been having some trouble with choosing the variables precision in my code... For a while i've been declaring variables as real(kind=8) :: var and I now understand that this is not very portable...
View ArticleClik here to view.
Inconsistent results with -fast flag
!bug.f90 program NN implicit none integer, parameter :: input_size = 2 integer, parameter :: hidden_size = 6 integer, parameter :: num_layer = 9 real*8 :: LN (hidden_size, hidden_size, num_layer)...
View Article