https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/Help-with-VS-project-config/m-p/1752853#M37617 <P>I'm using the MKL in C++ project: a Windows x64 DLL compiled with MSVC, and I wish to ship a single DLL with as fewer DLLs as possible.</P><OL><LI>I want to use the MKL via its NuGet package, instead of installing the whole toolkit. Which one should I use:&nbsp;intelmkl.devel.win-x64 or&nbsp;intelmkl.static.win-x64?</LI><LI>I get more than one categories in my project settings: "intelmkl.devel/static.win-x64", and "Intel (R) Performance Libraries". Which one (or both) should I edit?</LI><LI>I'm getting linker errors about runtime libraries mismatch on debug configuration, static link: object files compiled from my code is MDd_DynamicDebug, but mkl_tbb_thread.lib(bless_node_tbb.obj) is MD_DynamicRelease. Does the intelmkl.static.win-x64 package only has release libraries?</LI><LI>Help me choose between sequential and parallel (not cluster): If I'll be computing FFTs on 2000×2000 floating-point numbers, which one is better? If I need parallel, should I use TBB or OpenMP?</LI></OL> Wed, 01 Jul 2026 08:45:24 GMT candy76041820 2026-07-01T08:45:24Z https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/Help-with-VS-project-config/m-p/1752853#M37617 <P>I'm using the MKL in C++ project: a Windows x64 DLL compiled with MSVC, and I wish to ship a single DLL with as fewer DLLs as possible.</P><OL><LI>I want to use the MKL via its NuGet package, instead of installing the whole toolkit. Which one should I use:&nbsp;intelmkl.devel.win-x64 or&nbsp;intelmkl.static.win-x64?</LI><LI>I get more than one categories in my project settings: "intelmkl.devel/static.win-x64", and "Intel (R) Performance Libraries". Which one (or both) should I edit?</LI><LI>I'm getting linker errors about runtime libraries mismatch on debug configuration, static link: object files compiled from my code is MDd_DynamicDebug, but mkl_tbb_thread.lib(bless_node_tbb.obj) is MD_DynamicRelease. Does the intelmkl.static.win-x64 package only has release libraries?</LI><LI>Help me choose between sequential and parallel (not cluster): If I'll be computing FFTs on 2000×2000 floating-point numbers, which one is better? If I need parallel, should I use TBB or OpenMP?</LI></OL> Wed, 01 Jul 2026 08:45:24 GMT https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/Help-with-VS-project-config/m-p/1752853#M37617 candy76041820 2026-07-01T08:45:24Z