Meaning
Non-volatile storage devices designed for high-availability systems often divide their storage array into independent sectors to allow parallel operations. A dual-bank memory features two physically isolated regions that can be accessed simultaneously.
Hardware Partition
Silicon structures for such arrays include separate address decoders and charge pumps for each storage section. This isolation allows the microcontroller to execute critical firmware from the primary bank without encountering wait states or stalls during an erase operation on the secondary bank. Circuit layout ensures that the electrical stress of a write cycle does not bleed into the active reading area.
Update Execution
Firmware updates over the air require a highly resilient execution model to prevent device failure in the field. When a new image is received via the wireless transceiver, the application writes the update block by block into the inactive bank while continuing to run normally from the active bank. If a power failure occurs during this transfer, the running application remains fully intact and undamaged.
This architecture avoids the need for a secondary external storage chip on the printed circuit board.
System Availability
Recovery plans are simplified because the previous operational code is always preserved in the companion bank. This redundant structure eliminates the risk of bricking the hardware during standard maintenance schedules. Bootloader routines can easily verify the checksum of both banks and decide which one to execute, providing a seamless fallback option if the new version fails to boot.