Intel® C++ Compiler

Webinar on 9/27/2017 9am PST on Better Threaded Performance and Scalability with Intel Vtune Amplifer + OpenMP*

Webinar Topic: Better Threaded Performance and Scalability with Intel Vtune Amplifer + OpenMP*

Date: Wednesday, September 27, 2017 9 AM PDT

Abstract: Modern generation of Intel(R) Processors come with multiple processing cores and each core offering SIMD registers. This makes it all the more important for Application Developers to be aware of how to utilize all the processing cores efficiently. This webinar will walk step by step on the following concepts using a simple Edge Detection application:

Intel® C++ Compiler 18.0 has been released as part of Intel Parallel Studio XE 2018

Intel® Parallel Studio XE 2018, including Intel® C++ Compiler 18.0, is now available from the Intel Registration Center. Release notes can be found https://software.intel.com/en-us/articles/intel-c-compiler-180-release-notes.

Some of the new supported features are below:

Оптимизировали, оптимизировали, да не выоптимизировали!

Оптимизация? Конечно, каждый сталкивался с данной задачей при разработке своих, сколь-нибудь значительных, требующих определённых вычислений, приложений. При этом способов оптимизировать код существует огромное множество, и, как следствие, различных путей сделать это в автоматическом режиме с помощью опций компилятора. Вот здесь и возникает проблема – как выбрать то, что нужно нам и не запутаться?

best practice for overloading operators (&, |) for __int64 in icpc

Dear All,

One of our client (academic) researchers is using a third party R package which installs successfully on our compute cluster using R without MKL, gcc, and g++ (having support for said operators with __int64), but fails in R-with-MKL AFAIK due to lack of overload defined for said operators with __int64 in icpc.

I presume it's possible to introduce such overload definitions for use by icpc via additional, local code - independent of the said third party code, true?

If so, I'd appreciate if someone could provide a simple example or link thereto, thanks.

Intel C++ compiler not found by Visual Studio 2015

Hi,
I recently started working on a C++ project developed in visual studio. Once I downloaded the project from the repository and opened the solution in Visual Studio, I faced an error: C1083: Cannot open include file 'mkl.h': No such file or directory.

After downloading and installing Intel Math Kernal Library(and ergo Intel Parallel studio 17), the previous error was resolved. But now visual studio complains that it cannot find the Intel C++ compiler 17.0. 

Need AVX2 Intel v18 compile for slowest LZSS compressor paired with Fastest decompressor

Hope, someone compile Nakamichi with AVX2, wanna see how all incoming *lake CPUs behave in this AWESOME area.

After writing many variants of my amateurish simple LZSS compressor targeting textual data, here comes the strongest (with 1TB Sliding Window) - Nakamichi 'Ryuugan-ditto-1TB'.

Along with my wish to share the C source of this unique etude, I ask for AVX2 executable/compile with the latest Intel C optimizer.
You see, the attached file with the source and the Intel v15 SSE4.1 and AVX compiles has the compilation line:

Possible bug in C++ compiler of Intel compiler suite 2018

when I tried to install the PETSc 3.8.0 complex version I got

gmake[2]: *** [arch-linux2-c-opt/obj/src/mat/impls/baij/seq/baijsolvtrannat.o] Error 4
gmake[2]: *** Waiting for unfinished jobs....
/dev/shm/mathprod/PETSc/3.8.0/intel-para-2017b_complex/petsc-3.8.0/src/mat/impls/baij/seq/baijsolvtran.c(764) (col. 20): internal error: 04010002_0

Vectorization with OpenMPv4 and performance portability across compilers

Hi,

 

I am developping some vectorized code with OpenMP v4 and I am really happy with the performance I get with Intel compilers. Unfortunately, both gcc 7.2.0 and clang 5.0.0 make a mess of those (#pragma omp simd) and give very bad performance. This make the use of OpenMP v4 non portable across compilers in terms of performance. This makes writing software for AVX, AVX2 and AVX512 quite difficult as intrinsics seems to be the only solution if you want to get good performance across compilers.

Subscribe to Intel® C++ Compiler