The current version of Intel IPP supports the following data types of the source and destination for functions that perform matrix operations:
32f |
32-bit floating-point |
64f |
64-bit double precision. |
All objects of each particular IPP MX function must have the same type. Accordingly, it is assumed that sizes of the objects are specified in sizes of the data type, unless otherwise explicitly indicated.
Copyright © 2000 - 2011, Intel Corporation. All rights reserved.