|
|
|
|
|
|
by anonymous_iam
3413 days ago
|
|
|
I did a similar comparison last year, but mine was for an embedded system with no FPP (MSP430). I was able to shave more than half of the average execution time compared to the (highly optimized TI) math library function. The only drawback to my method was that it would fail for input values greater than 255. Fortunately my application was guaranteed to never supply an input value greater than 240. |
|
|