Memory module that includes a memory module copy engine for copying data from an active memory die to a spare memory die

1. A memory module comprising:

an array of memory dies arranged into logical rows and columns, with each memory die capable of assuming a status of spare, active, or retired;

system data, address, and clock connections for coupling system data, address, and clock signals to a system;

a number of memory module copy engines for copying data from a failing active memory die to a spare memory die, wherein each of the memory module copy engines is associated with each of the columns of the array of memory dies; and

an address/control unit for mapping the system data, address, and clock signals to memory data, address, and clock signals corresponding to active memory dies.

2. The memory module of claim 1 wherein all memory dies in a common row have a common status, and rows have the status of spare, active, or retired.

3. The memory module of claim 2 wherein the memory module copy engine copies data from an active row having a failing memory die to a spare row.

4. The memory module of claim 1 wherein at least two memory dies within each column of memory dies have a different status of the spare, active, or retired statuses.

5. A computer system comprising:

a processor;

persistent non-transitory storage coupled to the processor;

a memory controller either external to and coupled with the processor, or integrated within the processor, for generating system memory data, address, and clock signals; and

a memory module comprising:

an array of memory dies arranged into logical rows and columns, with each memory die capable of assuming a status of spare, active, or retired;

an address/control unit for mapping the system memory data, address, and clock signals from the memory controller to memory data, address, and clock signals corresponding to active memory dies; and

a number of memory module copy engines for copying data from a failing active memory die to a spare memory die, wherein each of the number of memory module copy engines use a source row register and a target row register to copy the data from the failing active memory die.

6. The computer system of claim 5 wherein the memory controller includes an error detection unit for detecting memory errors in the memory module, and upon reaching an error threshold, instructs at least one of the memory module copy engines to copy data from an active memory die that has reached the error threshold to a spare memory die, and instructs the address/control unit to change the status of the active memory die that has reached the error threshold from active to retired, and to change the status of the spare memory die that has received contents of the active memory die that has reached the error threshold from spare to active.

7. The computer system of claim 6 wherein all memory dies in a common row have a common status, rows have the status of spare, active, or retired, and the memory controller instructs the memory module copy engine to copy data from an active row having a memory die that has reached the error threshold to a spare row, and instructs the address/control unit to change the status of the active row containing the memory die that reached the error threshold from active to retired, and to change the status of the spare row that received the contents of the active row containing the memory die that has reached the error threshold from spare to active.

8. The computer system of claim 5 wherein at least two of the memory dies within each column of memory dies have a different status of the spare, active, or retired statuses.

9. The computer system of claim 5 wherein the memory controller includes a scrubbing unit, and upon copying data from a memory die that has reached an error threshold to a spare die, the scrubbing unit scrubs memory by applying error correction.

10. A method comprising:

detecting data errors in data stored in an active memory die;

instructing a number of memory module copy engines on a common memory module with the active memory die having the data errors to copy data from the active memory die having the data errors to a spare memory die, wherein:

each of the number of memory module copy engines is associated with one of the columns of the array of memory dies; and

the memory module copy engines copy the data using a source row register and a target row register;

mapping access from the active memory die having the data errors to the spare memory die; and

changing to retired a status of the active memory die having the data errors, and changing to active a status of the spare memory die receiving the data from the memory die having the data errors.

11. The method of claim 10 and further comprising:

scrubbing memory errors from the active memory die that was previously the spare memory die.

12. The method of claim 10 wherein all memory dies are organized into logical rows, rows have a status of spare, active, or retired, and the memory module copy engine copies data from an active row to a spare row.

13. The method of claim 10 wherein the memory dies are organized into columns, and each memory die within each column of memory dies has a status of spare, active, or retired, and a number of memory module copy engines are associated with each column of memory dies for copying data from a failing active memory die to a spare memory die within the column.

14. The method of claim 10 wherein the method elements are embodied on a non-transitory computer-readable medium.