I have two matricies (of type float *), so I guess they can be viewed as arrays.

I need to do an element-wise addition of the two matrices, and I would like to increase performance as much as possible. This will be for an iOS app, so I would like to see if the Accelerate.framework has such a function to do this.

The documentation for Accelerate.framework is a bit difficult to understand though. There are a bunch of functions with the word add in them, but I have no idea which or how they work.

Has anyone had to do this before? Can you point me in the correct direction?

NOTE: I will also need to do the same this for division (i.e., element-wise division); and I'd like to also use Accelerate.framework if possible.

有帮助吗?

解决方案

If your matrices can be viewed as arrays (are contiguous in memory), the functions you’re looking for are vDSP_vadd and vDSP_vdiv.

许可以下: CC-BY-SA归因
不隶属于 StackOverflow
scroll top