Currently, we use a combination of LPDDR4 memory for TDA4VM
As far as I know, the maximum bandwice of LPDDR4 is 17GB/s,
But, When using the "appPerfStatsDrStatsPrint" API supported by TI SDK7.1, the vlaue of PEAKs exceeds the maximum bandwidth(17GB/s).
Looking at the PEAK calculation part, it is as follows, but I would like to ask if it is the right way to measure by adding write and read as shown in the formula below.