AnsweredAssumed Answered

Why does AMCLIB multiple function run slower than not using it?

Question asked by Kidd Yang on Aug 27, 2018

I have tesed runtime of MLIB_Mul_FLT and float*float(e.g. 1.2f*1.3f), and once MLIB_Mul_FLT spends about 0.22us, but when I directly use 1.2f*1.3f, it only spends about 0.1us.

Since MLIB_Mul_FLT is a inlined function, shouldn't these two methods be the same?

Why does it happen?

By the way, is the runtime of floating point multiplication normal(0.1us once)?Or should I turn on MPC5744P's FPU first by some code?

Outcomes