Method and apparatus for memory fault tolerance
View Patent ↗One or more circuits may include an array of memory cells corresponding to a particular memory address. The one or more circuits may be operable to discover a location of a faulty memory cell in the array of memory cells. The one or more circuits may be operable to arrange the order in which the bits of a data block are stored to said array of memory cells based, at least in part, on said discovered location of said faulty memory cell.
1. A system comprising:
one or more circuits comprising an array of non-redundant memory cells corresponding to a particular memory address, said one or more circuits being operable to:
discover a location of a faulty memory cell in said array of memory cells; and
arrange the order in which the bits of a data block are stored to said array of memory cells based, at least in part, on said discovered location of said faulty memory cell, wherein said arrangement is such that:
when said discovered location is a first location, said bits of said data block are stored to said array of memory cells in descending order from most-significant bit to least-significant bit; and
when said discovered location is a second location, said bits of said data block are stored to said array of memory cells in ascending order from most-significant bit to least-significant bit.
2. The system of claim 1 , wherein:
said array of memory cells consists of N memory cells;
said data block consists of N data bits; and
N is an integer greater than 1.
3. The system of claim 1 , wherein said one or more circuits are operable to:
discover, for said data block to be stored to said array of memory cells, how sensitive said data block is to an error at a bit location of said data block corresponding to said location of said faulty memory cell in said array of memory cells; and
arrange the order in which said bits of said data block are stored to said array of memory cells based, at least in part, on said discovered sensitivity.
4. The system of claim 1 , wherein said one or more circuits are operable to process a datastream.
5. The system of claim 4 , wherein said processing of said datastream comprises deinterleaving said datastream.
6. The system of claim 1 , wherein said one or more circuits comprise a fault register that stores an indication of said discovered location of said faulty memory cell.
7. The system of claim 1 , wherein said one or more circuits comprise a multiplexer operable to perform said arrangement.
8. A system comprising:
one or more circuits comprising an array of non-redundant single-level memory cells, said one or more circuits being operable to:
arrange bits of data to be written to said array of non-redundant single-level memory cells based on a sensitivity coefficient assigned to one or more cells of said array of non-redundant single-level memory cells, wherein:
when said discovered location is a first location, said arranged bits of data are in descending order from most-significant bit to least-significant bit; and
when said discovered location is a second location, said arranged bits of data in ascending order from most-significant bit to least-significant bit; and
write said arranged bits of data to said array of non-redundant single-level memory cells.
9. The system of claim 8 , wherein a sensitivity coefficient of a cell of said array of non-redundant single-level memory cells is determined by comparing the value of a performance metric measured when said cell is functional to the value of said performance metric measured when said cell is faulty.
10. The system of claim 8 , wherein said one or more circuits comprise a multiplexer operable to perform said arrangement.