Addressing my original post again, the 27 spikes could simply be 27 data points that you entered at those specific rpms to form your data points. Taking the first derivative of discrete data points will create delta like spikes at those rpm points. They could coincide because you entered the data at the same rpm points for both charts. I am pretty sure it is an artifact of the data processing and not of actual engine operation. If they were real then the spikes will imply step like changes in hp at those rpm points, which I have never seen in a real dyno chart.