Reading electronic memory utilizing relationships between cell state distributions
View Patent ↗Providing distinction between overlapping state distributions of one or more multi cell memory devices is described herein. By way of example, a system can include a calculation component that can perform a mathematical operation on an identified, non-overlapped bit distribution and an overlapped bit distribution associated with the memory cell. Such mathematical operation can produce a resulting distribution that can facilitate identification by an analysis component of at least one overlapped bit distribution associated with cells of the one or more multi cell memory devices. Consequently, read errors associated with overlapped bits of a memory cell device can be mitigated.
1. A system that provides error correction for non-volatile memory, comprising:
a calculation component that mathematically operates on respective state distributions of a plurality of multi-bit memory cells, and outputs a resulting distribution of the mathematical operation that is different from each of the respective state distributions; and
an analysis component that employs the resulting distribution in correcting a bit error of a subset of the plurality of multi-bit memory cells.
2. The system of claim 1 , the plurality of multi-bit memory cells are mirror bit cells each comprising two or more adjacent transistors, and the respective state distributions are associated with respective adjacent transistors of the mirror bit cells.
3. The system of claim 1 , the calculation component selects the mathematical operation from a set of mathematical operations.
4. The system of claim 3 , the selection is based on a relationship between a position of a multi-bit cell on one of the respective state distributions with respect to a position of the multi-bit cell on another of the respective state distributions.
5. The system of claim 1 , the calculation component subtracts the respective state distributions to obtain the resulting distribution if a gate voltage or current level of a subset of the multi-bit memory cells is in a similar location on two or more of the respective state distributions.
6. The system of claim 1 , the calculation component adds the respective state distributions to obtain the resulting distribution if a gate voltage or current level of a subset of the multi-bit memory cells is in a dissimilar location on two or more of the respective state distributions.
7. The system of claim 6 , the dissimilar location corresponds to a high relative position on a first of the two or more respective state distributions and a low relative position on a second of the two or more respective state distributions.
8. The system of claim 7 , the first state distribution comprises bit levels of a first group of cells of the subset of the multi-bit memory cells and the second state distribution comprises bit levels of a second group of cells of the subset of the multi-bit memory cells.
9. The system of claim 8 , the subset of multi-bit memory cells are mirror bit cells comprising two or more adjacent transistors, and the first group of cells are adjacent to the second group of cells.
10. The system of claim 1 , a dispersity of the resulting distribution is smaller than respective dispersities of the respective state distributions.
11. The system of claim 1 , further comprising a reference component that generates a static current or voltage for identifying a state of at least one of the respective state distributions, the identified state and the resulting distribution enable correction of the bit error.
12. The system of claim 11 , further comprising a filtering component that disables memory cells associated with the at least one identified state distribution to facilitate correction of the bit error.
13. The system of claim 1 , further comprising:
a measurement component that determines a program state of each distinguishable state distribution of the respective state distributions; and
a program component that alters default state levels corresponding to one or more program or erase states of the non-volatile memory, to facilitate identifying a state of a non-distinguishable state distribution of the respective state distributions.
14. The system of claim 13 , wherein the program component at least one of:
changes default state levels of a program state overlapping the non-distinguishable state distribution; or
identifies a program state having minimal complementary bit disturbance and alters the default state levels of the one or more program or erase states closer to state levels of the program state having minimal disturbance.
15. The system of claim 13 , the analysis component determines at least one distinguishable state distribution in which a group of memory cells share a proportional or inverse proportional relationship with an adjacent group of memory cells of the non-distinguishable state distribution.
16. The system of 15 , the program component alters the default state levels of the one or more program or erase states such that the non-distinguishable state distribution overlaps the at least one distinguishable state distribution in which the group of memory cells share the relationship.
17. The system of claim 15 , the program component alters the default state levels of the one or more program or erase states to separate the non-distinguishable state distribution from a state distribution sharing no cell relationship with the non-distinguishable state distribution.
18. The system of claim 1 , the non-volatile memory comprises floating gate (FG) memory, multi-gate FG memory, charge trap (CT) memory, multi-gate CT dual-cell memory, NAND memory, AND memory or NOR memory, or a combination thereof.
19. A system that provides error correction for non-volatile memory, comprising:
means for mathematically combining values of respective state distributions of a plurality of multi-bit memory cells;
means for obtaining and storing a resulting distribution that is different from each of the respective state distributions and derived from combining the values thereof; and
means for employing the resulting distribution in correcting a bit error of a subset of the plurality of multi-bit memory cells.
20. A method of error correction for non-volatile memory, comprising:
employing electronic processing to execute device-executable instructions for resolving overlapping bit distributions of the non-volatile memory, the instructions comprising:
applying a mathematical relationship to values of respective state distributions of a plurality of multi-bit memory cells, wherein one or more values of the respective state distributions are logically linked by the mathematical relationship;
obtaining a resulting distribution from the mathematical relationship that is different from the respective state distributions; and
employing the resulting distribution to resolve at least one overlapping bit distribution of the non-volatile memory; and
employing memory to store the instructions.