Single precision float type Atomic function ?

Hi all

Atomic function can only do int type operation except atomicExch().
Is there any method to use atomic function to do single precision float
type operation? For example add.

Best wishes!

Usually floats aren’t accessed by atomics because due to roundoff and precision, FP addition is not communative, unlike integers.

You can roll your own with the exchange operator if you want though, there’s been several threads about it here, including one where I posted a code snippet a few months back.