Memory die remapping

1. A computer-implemented method, comprising:

detecting a trigger to configure a memory including a plurality of dice, each die of the plurality of dice accessed by a controller via one of a plurality of channels, each channel providing access to multiple dice, the plurality of dice including two spare dice, a first spare die of the two spare dice accessible via a first channel of the plurality of channels and a second spare die of the two spare dice accessible via a second channel of the plurality of channels; and

in response to detecting the trigger, mapping a plurality of logical units to the plurality of dice, excluding the two spare dice, wherein the mapping includes mapping each logical unit of the plurality of logical units across multiple dice of the plurality of dice such that:

a first set of the plurality of logical units reside on dice accessible via a first group of channels including the second channel and excluding the first channel,

a second set of the plurality of logical units reside on dice accessible via a second group of channels including the first channel and excluding the second channel,

the first spare die is a spare for the second set of logical units such that remapping a die of the plurality of dice mapped to the second logical units to the first spare die avoids channel stacking, and

the second spare die is a spare for the first set of logical units such that remapping a die of the plurality of dice mapped to the first logical units to the second spare die avoids channel stacking.

2. The computer-implemented method of claim 1 , further comprising:

performing an error analysis of the plurality of dice;

determining, using the error analysis, two dice having higher error rates than others of the plurality of dice; and

remapping logical units that were mapped to the two dice having higher error rates to the spare dice.

3. The computer-implemented method of claim 2 , wherein the remapping includes writing die addresses of the two dice having higher error rates to a data structure corresponding to spare dice.

4. The computer-implemented method of claim 3 , further comprising:

detecting an operation targeted at a portion of a logical unit;

determining a die address of a die targeted by the operation matches one of the die addresses written to the data structure; and

performing the operation using the die address of the spare die to which that entry in the data structure corresponds.

5. The computer-implemented method of claim 2 , wherein determining two dice having higher error rates includes determining a first die having a highest error rate of dice mapped to the first set of the plurality of logical units and a second die having a highest error rate of dice mapped to the second set of the plurality of logical units, and wherein the remapping includes remapping a portion of a logical unit that was mapped to the first die having a highest error rate to the first spare die and remapping a portion of a logical unit that was mapped to the second die having a highest error rate to the second spare die.

6. The computer-implemented method of claim 1 , wherein each logical unit is mapped to a group of physically contiguous dice.

7. The computer-implemented method of claim 1 , wherein memory capacity of the two spare dice is omitted from memory capacity reported to a user.

8. A non-transitory computer-readable medium storing instructions, which when executed by a processing device, cause the processing device to perform a method comprising:

computer-implemented method, comprising:

detecting a trigger to configure a memory including a plurality of dice, each die of the plurality of dice accessed by a controller via one of a plurality of channels, each channel providing access to multiple dice, the plurality of dice including two spare dice, a first spare die of the two spare dice accessible via a first channel of the plurality of channels and a second spare die of the two spare dice accessible via a second channel of the plurality of channels; and

in response to detecting the trigger, mapping a plurality of logical units to the plurality of dice, excluding the two spare dice, wherein the mapping includes mapping each logical unit of the plurality of logical units across multiple dice of the plurality of dice such that:

a first set of the plurality of logical units reside on dice accessible via a first group of channels including the second channel and excluding the first channel,

a second set of the plurality of logical units reside on dice accessible via a second group of channels including the first channel and excluding the second channel,

the first spare die is a spare for the second set of logical units such that remapping a die of the plurality of dice mapped to the second logical units to the first spare die avoids channel stacking, and

the second spare die is a spare for the first set of logical units such that remapping a die of the plurality of dice mapped to the first logical units to the second spare die avoids channel stacking.

9. The non-transitory computer-readable medium of claim 8 , the method further comprising:

performing an error analysis of the plurality of dice;

determining, using the error analysis, two dice having higher error rates than others of the plurality of dice; and

remapping logical units that were mapped to the two dice having higher error rates to the spare dice.

10. The non-transitory computer-readable medium of claim 9 , wherein the remapping includes writing die addresses of the two dice having higher error rates to a data structure corresponding to spare dice.

11. The non-transitory computer-readable medium of claim 10 , the method further comprising:

detecting an operation targeted at a portion of a logical unit;

determining a die address of a die targeted by the operation matches one of the die addresses written to the data structure; and

performing the operation using the die address of the spare die to which that entry in the data structure corresponds.

12. The non-transitory computer-readable medium of claim 9 , wherein determining two dice having higher error rates includes determining a first die having a highest error rate of dice mapped to the first set of the plurality of logical units and a second die having a highest error rate of dice mapped to the second set of the plurality of logical units, and wherein the remapping includes remapping a portion of a logical unit that was mapped to the first die having a highest error rate to the first spare die and remapping a portion of a logical unit that was mapped to the second die having a highest error rate to the second spare die.

13. The non-transitory computer-readable medium of claim 8 , wherein each logical unit is mapped to a group of physically contiguous dice.

14. The non-transitory computer-readable medium of claim 8 , wherein memory capacity of the two spare dice is omitted from memory capacity reported to a user.

15. A system comprising:

a plurality of dice including two spare dice; and

a processing device coupled to the plurality of dice via a plurality of channels, each channel providing access to multiple dice, wherein a first spare die of the two spare dice is accessible via a first channel of the plurality of channels and a second spare die of the two spare dice is accessible via a second channel of the plurality of channels, wherein the processing device executes instructions to:

detect a trigger to configure the system; and

in response to detecting the trigger, map a plurality of logical units to the plurality of dice, excluding the two spare dice, wherein the mapping includes mapping each logical unit of the plurality of logical units across multiple dice of the plurality of dice such that:

a first set of the plurality of logical units reside on dice accessible via a first group of channels including the second channel and excluding the first channel,

a second set of the plurality of logical units reside on dice accessible via a second group of channels including the first channel and excluding the second channel,

the first spare die is a spare for the second set of logical units such that remapping a die of the plurality of dice mapped to the second logical units to the first spare die avoids channel stacking, and

the second spare die is a spare for the first set of logical units such that remapping a die of the plurality of dice mapped to the first logical units to the second spare die avoids channel stacking.

16. The system of claim 15 , wherein the instructions further cause the processing device to:

perform an error analysis of the plurality of dice;

determine, using the error analysis, two dice having higher error rates than others of the plurality of dice; and

remap logical units that were mapped to the two dice having higher error rates to the spare dice.

17. The system of claim 16 , wherein the remapping includes writing die addresses of the two dice having higher error rates to a data structure corresponding to spare dice.

18. The system of claim 17 , wherein the instructions further cause the processing device to:

detect an operation targeted at a portion of a logical unit;

determine a die address of a die targeted by the operation matches one of the die addresses written to the data structure; and

perform the operation using the die address of the spare die to which that entry in the data structure corresponds.

19. The system of claim 16 , wherein determining two dice having higher error rates includes determining a first die having a highest error rate of dice mapped to the first set of the plurality of logical units and a second die having a highest error rate of dice mapped to the second set of the plurality of logical units, and wherein the remapping includes remapping a portion of a logical unit that was mapped to the first die having a highest error rate to the first spare die and remapping a portion of a logical unit that was mapped to the second die having a highest error rate to the second spare die.

20. The system of claim 15 , wherein each logical unit is mapped to a group of physically contiguous dice.