Based on the time for feature extraction and prediction for each wake word, it was determined that the Raspberry Pi 4 with 4 GB of RAM is the current optimal solution. Each wake word buffer takes about 128 ms per prediction so the CPU will need to accomplish each prediction in less than that time.