Why FMD need built in battery ?
All of these solid state storage devices have some sort of microprocessor and volatile memory in them. In the case of an FMD, the volatile memory is DDR3. This memory is used for writes in flight, caching, and most importantly the "Flash Translation Table". The Flash Translation Table is where the FMD (or SSD) maps LBAs to physical memory locations in the flash memory space. If you lose this, you lose your data. As a side note, in the early days of SSDs, the idea that "SSDs lose your data" was because early FW would end up corrupting the Flash Translation Table.
In the event of an unanticipated power down - some ugly power loss or someone pulls the wrong drive during maintenance - the Flash Translation Table must be preserved. So the microprocessor needs enough guaranteed power to run long enough to back up the Flash Translation Table, in this case to flash.
Different SSD designers handle the backing up of the Flash Translation Table in different ways. Some just keep it in flash, but this results in lower overall performance. Some build a capacitor-based solution using supercaps or electrolytic capacitors. If the microprocessor's power requirements are low enough, this can be viable, but the energy storage density for a given capacity of capacitor solution is not that dense as a battery. Hence the reason FMDs use a battery.
We have a pretty powerful microprocessor in the FMD's ASIC and, because the capacities of FMDs are bigger than traditional SSDs, the Flash Translation Table is bigger, so we need the extra juice the battery provides to ensure that data is never compromised in a power down situation.
When I looked at these drives with an engineer it appeared the battery would be used to drain the volatile internal DRAM in the event of a sudden power loss.
“The Internal battery is used to ensure that data is saved from the internal DRAM in the FMD to the flash memory in the event of an unscheduled power outage”
Retrieving data ...