Due to their limited resources, IoT devices are constantly striving to find a balance between energy consumption and performance. In this ongoing endeavor, the communication module frequently emerges as the foremost contributor to energy consumption. Among various communication technologies, Bluetooth Low Energy (BLE) stands out as a promising candidate due to its lightweight architecture and excellent performance. Despite recent advances in BLE optimization, the Neighbor Discovery Process (NDP) remains a highly energy-cost task. To enhance energy efficiency during this operation, we propose a new approach for the NDP that leverages the multicore architecture of a recent microcontroller to dynamically adjust the scanning duration based on the number of devices to discover. Intensive simulations have shown that a fixed scanning duration can be avoided by ending the NDP after discovering the required number of devices. Our proposal demonstrates a reduction in energy consumption through intensive simulations and a robust validation methodology. Furthermore, real-world experiments indicate that our approach can achieve up to 50% energy savings compared to a 1-second scan when only one core is used during the NDP.