Programmable Devices
CPLDs, FPGAs, SoC FPGAs, Configuration, and Transceivers
21602 Discussions

Matrix Inverse IP core for Stratix V FPGA

Altera_Forum
Honored Contributor II
‎07-28-2015 10:54 PM
1,846 Views

Does the input to the Floating Point Matrix Inverse IP core require a square matrix as its inverse? I want to get a 12x12 inverse using the Cholesky decomposition by inputting a 12x200 matrix. Can this be done with the IP core?

0 Kudos

Link Copied

1 Reply
Altera_Forum
Honored Contributor II
‎08-03-2015 11:29 PM
789 Views

I don't think it is supported, based on the datasheet (https://www.altera.com/content/dam/altera-www/global/en_us/pdfs/literature/ug/ug_altfp_mfug.pdf) for the IP: 

 

"DIMENSION: The number of rows in the matrix. As the matrix is 

square, this is also the number of columns in the matrix. 

The supported dimensions are 4 x 4, 6 x 6, 8 x 8, 16 x 16, 

32 x 32, and 64 x 64." 

 

It seems to imply that a square matrix is required. And 200x12 is larger than the 64 maximum anyway. 

 

Besides, I wasn't aware you can find an inverse for a non-square matrix. Been quite some time since I did matrix stuff in detail, so I could quite easily be wrong on that.
0 Kudos
Copy link
Reply

Community support is provided Monday to Friday. Other contact methods are available here.

Intel does not verify all solutions, including but not limited to any file transfers that may appear in this community. Accordingly, Intel disclaims all express and implied warranties, including without limitation, the implied warranties of merchantability, fitness for a particular purpose, and non-infringement, as well as any warranty arising from course of performance, course of dealing, or usage in trade.

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