Data storage device and method for data routing for wear levelling in multi-meta-die data storage devices
A multi-meta-die data storage device disclosed herein routes host writes in such a way to prevent the need for inter-meta-die active wear leveling. In making the meta-die selection, a controller of the data storage device can consider both the average number of program-erase cycles (PEC) of the meta-die, as well as a parameter that indicates the effort required to relocate data from the meta-die to another meta-die. The controller can use this information to normalize PEC among the meta-dies and select the meta-die with the lowest final relocation effort for opening a new block for host writes. This can minimize relocations from active wear leveling.
1. A data storage device comprising:
a memory comprising a plurality of memory dies organized into a plurality of meta-dies, wherein each meta-die comprises two or more memory dies that operate in parallel; and
one or more processors, individually or in combination, configured to:
receive data from a host to be written in the memory, wherein writing the data in the memory requires a new block to be opened in the memory;
determine which meta-die from the plurality of meta-dies in which to open the new block to write the data received from the host by:
determining an average number of program-erase cycles (PEC) of each of the plurality of meta-dies;
for each meta-die, determining an effort required to relocate data from that meta-die to another meta-die to free up space to create the new block; and
based on the average number of program-erase cycles of each of the plurality of meta-dies and the effort required to relocate data from each of the plurality of meta-dies, selecting a meta-die from the plurality of meta-dies in which to open the new block to write data received from the host; and
write data received from the host in the new block.
2. The data storage device of claim 1 , wherein the selecting is further based on a mode of the data storage device.
3. The data storage device of claim 2 , wherein the mode comprises a burst mode, a sustained garbage collection mode, an urgent garbage collection mode, and/or a super-urgent garbage collection mode.
4. The data storage device of claim 1 , wherein the effort required to relocate data from each of the plurality of meta-dies is determined by an average validity count of relocation sources to free up an equivalent block.
5. The data storage device of claim 1 , wherein the effort required to relocate data from each of the plurality of meta-dies is determined using a look-up table.
6. The data storage device of claim 1 , wherein among the plurality of meta-dies, the selected meta-die is least-likely to require inter-meta-die active wear leveling.
7. The data storage device of claim 1 , wherein each meta-die comprises a maximum number of memory dies that can operate in parallel.
8. The data storage device of claim 1 , wherein each meta-die is an independent sub-system with respect to the other meta-dies of the plurality of meta-dies.
9. The data storage device of claim 1 , wherein each meta-die comprises its own group allocation table (GAT) blocks, host blocks, and/or relocation blocks.
10. The data storage device of claim 1 , wherein the memory comprises a three-dimensional memory.
11. In a data storage device comprising a memory comprising a plurality of memory dies organized into a plurality of meta-dies, wherein each meta-die comprises two or more memory dies that operate in parallel, a method comprising:
receiving data from a host to be written in the memory, wherein writing the data in the memory requires a new block to be opened in the memory;
determining which meta-die from the plurality of meta-dies in which to open the new block to write the data received from the host by:
determining an average number of program-erase cycles (PEC) of each of the plurality of meta-dies;
for each meta-die, determining an effort required to relocate data from that meta-die to another meta-die to free up space to create the new block; and
based on the average number of program-erase cycles of each of the plurality of meta-dies and the effort required to relocate data from each of the plurality of meta-dies, selecting a meta-die from the plurality of meta-dies in which to open the new block to write the data received from the host; and
writing the data received from the host in the new block.
12. The method of claim 11 , wherein the effort required to relocate data from each of the plurality of meta-dies is determined by an average validity count.
13. The method of claim 11 , wherein the meta-die is selected by considering a mode of the data storage device.
14. The method of claim 13 , wherein the mode comprises a burst mode, a sustained garbage collection mode, an urgent garbage collection mode, and/or a super-urgent garbage collection mode.
15. The method of claim 11 , wherein each meta-die comprises a maximum number of memory dies that can operate in parallel.
16. The method of claim 11 , wherein each meta-die is an independent sub-system with respect to the other meta-dies of the plurality of meta-dies.
17. The method of claim 11 , wherein the memory comprises a three-dimensional memory.
18. A data storage device comprising:
a memory comprising a plurality of memory dies organized into a plurality of meta-dies, wherein each meta-die comprises two or more memory dies that operate in parallel; and
means for:
receiving data from a host to be written in the memory, wherein writing the data in the memory requires a new block to be opened in the memory;
determining which meta-die from the plurality of meta-dies in which to open the new block to write the data received from the host by:
determining an average number of program-erase cycles (PEC) of each of the plurality of meta-dies;
for each meta-die, determining an effort required to relocate data from that meta-die to another meta-die to free up space to create the new block; and
based on the average number of program-erase cycles of each of the plurality of meta-dies and the effort required to relocate data from each of the plurality of meta-dies, selecting a meta-die from the plurality of meta-dies in which to open the new block to write the data received from the host; and
writing the data received from the host in the new block.