https://community.intel.com/t5/Intel-Integrated-Performance/Speed-of-function-in-IPP/m-p/865879#M8326 <P>All work is done inside DLL except data marshalling required by managed runtime.</P> <P>Vladimir</P> Mon, 30 Jun 2008 17:53:01 GMT Vladimir_Dudnik 2008-06-30T17:53:01Z https://community.intel.com/t5/Intel-Integrated-Performance/Speed-of-function-in-IPP/m-p/865878#M8325 <P>Hello. I am using IPP library with .NET. I do it somehow like this: <BR />[DllImport(lib)]<BR />public static extern IppStatus ippsCrossCorr_32fc(Complex_f* pSrc1, int len1, Complex_f* pSrc2, int len2, Complex_f* pDst, int dstLen, int lowLag);</P> <P>I tested the speed of this function's execution in C++ and C# and C++ is about 10% better. How can you explain that? I know that C# is managed and a bit slower than unmanaged C++ but there is no reason to be slow in this case because all work is done inside binary dll. Thanks.</P> Mon, 30 Jun 2008 15:59:29 GMT https://community.intel.com/t5/Intel-Integrated-Performance/Speed-of-function-in-IPP/m-p/865878#M8325 postaquestion 2008-06-30T15:59:29Z https://community.intel.com/t5/Intel-Integrated-Performance/Speed-of-function-in-IPP/m-p/865879#M8326 <P>All work is done inside DLL except data marshalling required by managed runtime.</P> <P>Vladimir</P> Mon, 30 Jun 2008 17:53:01 GMT https://community.intel.com/t5/Intel-Integrated-Performance/Speed-of-function-in-IPP/m-p/865879#M8326 Vladimir_Dudnik 2008-06-30T17:53:01Z