Computes the dot product of two real vectors.
event dot(queue &exec_queue, std::int64_t n, const T *x, std::int64_t incx, const T *y, std::int64_t incy, T_res *result, const vector_class<event> &dependencies = {});
The USM version of dot supports the following precisions and devices.
T | T_res | Devices Supported |
---|---|---|
float | float | Host, CPU, and GPU |
double | double | Host, CPU, and GPU |
float | double | Host, CPU, and GPU |
The dot routines perform a dot product between two vectors:
For the mixed precision version (inputs are float while result is double), the dot product is computed with double precision.
The queue where the routine should be executed.
Number of elements in vectors x and y.
Pointer to the input vector x. The array holding the vector x must be of size at least (1 + (n – 1)*abs(incx)). See Matrix and Vector Storage for more details.
Stride of vector x.
Pointer to the input vector y. The array holding the vector y must be of size at least (1 + (n – 1)*abs(incy)). See Matrix and Vector Storage for more details.
Stride of vector y.
List of events to wait for before starting computation, if any. If omitted, defaults to no dependencies.
Pointer to where the result (a scalar) will be stored.
Output event to wait on to ensure computation is complete.