hello..
i'm running a few lines of code and noticed the execution of the floating point math is pretty slow.. costing 10us to do one integer x float.. i'm using the DS for S32 power R17.1 and the Target Processor setting page of the property has the "Use hardware floating point" checked.. the user manual says single precision is fully supported but didn't elaborate the performance part. Any one any comments?
thanks
gzhang
solved... add suffix F or f to constants so that they are deemed as single precision. otherwise they are considered double.. takes a lot longer.
For this way, how much time do you take for float multiplication. Could you tell me?