Developer Reference for Intel® oneAPI Math Kernel Library - C
Vector Arguments in Vector Math
Vector arguments of classic VM mathematical functions are passed in one-dimensional arrays with unit vector increment. It means that a vector of length
is passed contiguously in an array
whose values are defined as
a, a, ..., a[
Strided VM mathematical functions allow using positive increments for all input and output vector arguments.
To accommodate for arrays with other increments, or more complicated indexing, VM contains auxiliary Pack/Unpack functions that gather the array elements into a contiguous vector and then scatter them after the computation is complete.
Generally, if the vector elements are stored in a one-dimensional array
a[m0], a[m1], ..., a[mn-1]
and need to be regrouped into an array
y[k0], y[k1], ..., y[kn-1],
VM Pack/Unpack functions can use one of the following indexing methods:
Positive Increment Indexing
kj = incy * j, mj = inca * j, j = 0 ,..., n-1
incy > 0
inca > 0
For example, setting incy = 1 specifies gathering array elements into a contiguous vector.
This method is similar to that used in BLAS, with the exception that negative and zero increments are not permitted.
Index Vector Indexing
kj = iy[j], mj = ia[j], j = 0 ,..., n-1
where ia and iy are arrays of length
that contain index vectors for the input and output arrays
Mask Vector Indexing
are such that:
0 , j = 0,..., n-1
are arrays that contain mask vectors for the input and output arrays