Draft: Added scripts to calculate flops

This MR adds two scripts that calculate the FLOPS, MIPS and GB/s of a given Allen sequence.

Usage: from the build directory

../scripts/profile_kernels.sh <sequence>

Example output:

Profile data:
Half precision ops one iteration: 651916571.0
Single precision ops one iteration: 696758718.0
Double precision ops one iteration: 0.0
Bytes read one iteration: 311826330.0
Bytes written one iteration: 270228430.0
Number of instructions one iteration: 506673481.0
Number of iterations: 3200.0
Time: 9.270162 s

Indicators:
Half precision TFLOPS: 0.2250373863153632
Single precision TFLOPS: 0.2405166055997727
Mixed half/single precision TFLOPS: 0.4655539919151359
Double precision TFLOPS: 0.0
Bytes read in GBps: 107.64043346815298
Bytes write in GBps: 93.2811072773054
MIPS: 174900.4104998381
IPC: 0.29158040728075246