Showing results for

- Intel Community
- Software Development SDKs and Libraries
- Intel® Integrated Performance Primitives
- which libraries to include in matrix multiplication program

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Mute
- Printer Friendly Page

Highlighted
##

Well,I am writing a matrix matrix multiplication program using ipp libraries but i dont know which libraries are to be linked at runtime .Please help

prakrati

Beginner

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

08-23-2010
01:01 AM

24 Views

which libraries to include in matrix multiplication program

3 Replies

Highlighted
##

Hello,

thanks for your interest to Intel IPP product.

I would recomend you to check IPP Reference Manual, volume 3, Small Matrices and Realistic Rendering, for more details on relevant functions.

For example, Multiply matrix-matrix operations supported in IPP are(you may find more operations in IPP document):

IPP library you need to link with is ippm.lib (and header file is ippm.h).

Please also note that in IPP we provide optimization for special case of small matrix sizes (like 3x3, 5x5, 7x7, 9x9, 11x11). In case you work with matrices of bigger size I would recomend you to take a look at Intel Math Kernel library (Intel MKL), which provides optimization for big matrices.

Regards,

Vladimir

Vladimir_D_Intel

Employee

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

08-23-2010
02:49 AM

24 Views

thanks for your interest to Intel IPP product.

I would recomend you to check IPP Reference Manual, volume 3, Small Matrices and Realistic Rendering, for more details on relevant functions.

For example, Multiply matrix-matrix operations supported in IPP are(you may find more operations in IPP document):

**Case 13: Matrix - matrix operation**

IppStatus ippmMul_mm_32f(const Ipp32f* pSrc1, int src1Stride1, int src1Stride2, int src1Width, int src1Height, const Ipp32f* pSrc2, int src2Stride1, int src2Stride2, int src2Width, int src2Height, Ipp32f* pDst, int dstStride1, int dstStride2);

IppStatus ippmMul_mm_64f(const Ipp64f* pSrc1, int src1Stride1, int src1Stride2, int src1Width, int src1Height, const Ipp64f* pSrc2, int src2Stride1, int src2Stride2, int src2Width, int src2Height, Ipp64f* pDst, int dstStride1, int dstStride2);

IppStatus ippmMul_mm_32f_P(const Ipp32f** ppSrc1, int src1RoiShift, int src1Width, int src1Height, const Ipp32f** ppSrc2, int src2RoiShift, int src2Width, int src2Height, Ipp32f** ppDst, int dstRoiShift);

IppStatus ippmMul_mm_64f_P(const Ipp64f** ppSrc1, int src1RoiShift, int src1Width, int src1Height, const Ipp64f** ppSrc2, int src2RoiShift, int src2Width, int src2Height, Ipp64f** ppDst, int dstRoiShift);

Please also note that in IPP we provide optimization for special case of small matrix sizes (like 3x3, 5x5, 7x7, 9x9, 11x11). In case you work with matrices of bigger size I would recomend you to take a look at Intel Math Kernel library (Intel MKL), which provides optimization for big matrices.

Regards,

Vladimir

Highlighted
##

prakrati,

Gennady_F_Intel

Moderator

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

08-23-2010
04:44 AM

24 Views

prakrati,

could you please look at this KB article:"Building apllication Using IPP v 5* and v.6*". You can find there the list of libraries you need tol link depend of linking mode...

--Gennady

Highlighted
##

forget to add, in the case if you use IPP 7.0 beta version, please look at this KB, which will help you to understand which libraries you will need to link with 7.0 beta.

Gennady_F_Intel

Moderator

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

08-23-2010
04:52 AM

24 Views

For more complete information about compiler optimizations, see our Optimization Notice.