DIVISION OPERATIONS ON VARIABLE LENGTH ELEMENTS IN MEMORY

• US 20160266873A1
• Filed: 03/08/2016
• Published: 09/15/2016
• Est. Priority Date: 03/11/2015
• Status: Active Grant

• Current Assignee: Micron Technology, Inc.
• Original Assignee: Micron Technology, Inc.
• Inventors: Tiwari, Sanjay, Wheeler, Kyle B.
• Granted Patent: US 9,898,253 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 2207/535 Indexing scheme relating to groups G06F7/535 - G06F7/5375
• CPC Class Codes: G06F 7/535 Dividing only
• CPC Class Codes: G11C 11/4076 Timing circuits for regeneration management G11C11/406
• CPC Class Codes: G11C 11/4091 Sense or sense/refresh amplifiers, or associated sense circuitry, e.g. for coupled bit-line precharging, equalising or isolating
• CPC Class Codes: G11C 7/06 Sense amplifiers; Associated circuits , e.g. timing or triggering circuits
• CPC Class Codes: G11C 7/10 Input/output [I/O] data interface arrangements, e.g. I/O data control circuits, I/O data buffers
• CPC Class Codes: G11C 7/1006 Data managing, e.g. manipulating data before writing or reading out, data bus switches or control circuits therefor

Abstract

Examples of the present disclosure provide apparatuses and methods for performing variable bit-length division operations in a memory. An example method comprises performing a variable length division operation on a first vector comprising variable length elements representing a number of dividends and stored in a group of memory cells coupled to a first access line and a number of sense lines of a memory array and a second vector comprising variable length elements representing a number of divisors stored in a group of memory cells coupled to a second access line and the number of sense lines of the memory array. The method can include dividing the first vector by the second vector by performing a number of operations. The method can include performing at least one of the number of operations without transferring data via an input/output (I/O) line.

Claims

1. A method for performing division operations comprising:

performing a division operation on;

a first vector comprising at least two elements that are different in length representing a number of dividends and initially stored in a group of memory cells coupled to a first access line and a number of sense lines of a memory array; and

a second vector comprising at least two elements that are different in length representing a number of divisors and initially stored in a group of memory cells coupled to a second access line and the number of sense lines of the memory array;

wherein the division operation includes dividing the first vector by the second vector by performing a number of operations and at least one of the number of operations are performed without transferring data via an input/output (I/O) line.

2. The method of claim 1, wherein performing at least one of the number of operations without transferring data via an I/O line comprises performing a number of AND operations, OR operations, and INVERT operations without a sense line address access.

3. The method of claim 1, wherein a result of the division operation is stored in a group of memory cells coupled to a third access line.

4. The method of claim 3, wherein the result of the division operation is stored in at least one of the group of memory cells coupled to the first access line and the group of memory cells coupled to the second access line.

5. The method of claim 1, wherein the number of operations are performed using sensing circuitry coupled to each of a number of columns of complementary sense lines.

6. The method of claim 5, wherein each of the sensing circuitry comprises a sense amplifier and a compute component.

7. The method of claim 6, wherein the sense amplifier comprises a primary latch and the compute component comprises a secondary latch.

8. The method of claim 1, further comprising storing a first element and a second element of the first vector in a first group of memory cells coupled to the first access line, wherein each memory cell in the first group of memory cells stores a respective bit of the first vector.

9. The method of claim 8, further comprising storing a third element and a fourth element of the second vector in a second group of memory cells coupled to the second access line, wherein each memory cell in the second group of memory cells stores a respective bit of the second vector.

10. The method of claim 9, wherein the first element and the third element have a first length and the second element and the fourth element have a second length.

11. An apparatus comprising:

a first group of memory cells in a memory array coupled to a first access line and configured to store a first dividend element;

a second group of memory cells coupled to the first access line and configured to store a second dividend element;

a third group of memory cells coupled to a second access line and configured to store a first divisor element;

a fourth group of memory cells coupled to the second access line and configured to store a second divisor element; and

a controller configured to operate sensing circuitry to divide the first dividend element by the first divisor element and to divide the second dividend element by the second divisor element by performing a number of operations, wherein at least one of the number of operations is performed without transferring data via an input/output (I/O) line.

12. The apparatus of claim 11, wherein the first group of memory cells and the third group of memory cells each comprise a same X number of memory cells and the second group of memory cells and the fourth group of memory cells each comprise a same Y number of memory cells.

13. The apparatus of claim 11, wherein the controller is configured to operate the sensing circuitry to divide the first dividend element by the first divisor element in parallel with dividing the second dividend element by the second divisor element.

14. The apparatus of claim 11, wherein the controller is configured to operate the sensing circuitry to perform a first variable compare operation with the first dividend element and the first divisor element by performing a number of AND operations, OR operations, and INVERT operations without transferring data via an input/output (I/O) line.

15. The apparatus of claim 14, wherein the controller is configured to operate the sensing circuitry to perform a second variable compare operation with the second dividend element and the second divisor element by performing a number of AND operations, OR operations, and INVERT operations without transferring data via an input/output (I/O) line;

and

wherein the first compare operation and the second compare operation are performed in parallel.

16. The apparatus of claim 15, wherein:

a result of the first variable compare operation is an indication of whether the first dividend element is less than, equal to, or greater than the first divisor element; and

a result of the second variable compare operation is an indication of whether the second dividend element is less than, equal to, or greater than the first divisor element.

17. The apparatus of claim 14, wherein the controller is configured to operate the sensing circuitry to perform the first variable compare operation comprises the sensing circuitry configured to:

create a first destination bit-vector that identifies a difference of the first dividend element as compared to the first divisor element;

create a second destination bit-vector that identifies a difference in the first divisor element as compared to the first dividend element.

18. The apparatus of claim 17, wherein the first destination bit-vector and the second destination bit-vector are the results of the first variable compare operation.

19. The apparatus of claim 17, wherein the controller is configured to operate the sensing circuitry to:

replicate bits in the first destination bit-vector that identify the difference in the first dividend element as compared to the first divisor element in a first replication phase; and

replicate bits in the second destination bit-vector that identify the difference in the first divisor element as compared to the first dividend element in the first replication phase.

20. The apparatus of claim 19, wherein the controller is configured to operate the sensing circuitry to:

store in the first destination bit-vector the difference in the first dividend element as compared to the first divisor element; and

store in the second destination bit-vector a difference in the first divisor element as compared to the first dividend element.

21. The apparatus of claim 20, wherein the controller is configured to operate the sensing circuitry to:

replicate bits in the first destination bit-vector in a second replication phase; and

replicate bits in the second destination bit-vector in the second replication phase;

wherein replicating the bits in the first replication phase and the second replication phase includes performing a number of right SHIFT operations.

22. The apparatus of claim 21, wherein the controller is configured to operate the sensing circuitry to store a static mask that defines a vector length for each of the variable length bit-vectors.

23. The apparatus of claim 22, wherein a quantity of the number of SHIFT operations performed is based on a longest vector length from among the vector lengths of each of the variable length bit-vectors.

24. The apparatus of claim 11, wherein:

the first dividend element and the first divisor element are a first bit-length;

the second dividend element and the second divisor element are a second bit-length; and

the first bit-length and the second bit-length are different bit-lengths.

25. A method for performing a number of division operations on variable length elements, comprising:

using a controller configured to operate sensing circuitry for performing a division operation on;

a first vector comprising at least two elements that are different in length representing a number of dividends and initially stored in a group of memory cells coupled to a first access line and a number of sense lines of a memory array;

a second vector comprising at least two elements that are different in length representing a number of divisors and initially stored in a group of memory cells coupled to a second access line and the number of sense lines of the memory array;

wherein the division operation includes dividing the first vector by the second vector by performing a number of operations on pitch with the memory array; and

storing a result of the division operation as a third vector stored in a group of memory cells coupled to a third access line.

26. The method of claim 25, wherein performing the division operation comprises performing a number of iterations of operations using:

a subtraction mask bit-vector;

a zero dividend bit-vector;

a quotient bit-vector; and

a remainder bit-vector;

wherein each iteration of operations comprises determining whether a remainder value represented by the remainder element remains to be divided.

27. The method of claim 26 determining a portion of a number of dividends divisible by a number of divisors and a remainder for the portion.

28. The method of claim 27, wherein the first iteration of the number of iterations of operations is performed by the controller configured to operate the sensing circuitry to:

compare elements of the first vector with the variable length elements of the second vector that results in the subtraction mask bit-vector;

subtract elements of the second vector from the variable length elements of the first vector using the subtraction mask bit-vector; and

store the result of the subtraction as the remainder bit-vector.

29. The method of claim 28, wherein the first iteration is performed by the controller configured to further operate the sensing circuitry to:

increment the quotient bit-vector by a count of one corresponding to the subtraction mask bit-vector;

perform an XOR operation on the remainder bit-vector and the first bit-vector; and

store a result of the XOR operation in a number of corresponding compute components of the sensing circuitry.

30. The method of claim 29, wherein a BlockOR operation is performed on the result of the XOR operation stored in the number of corresponding compute components by determining if the result includes a particular value.

31. The method of claim 30, wherein when the result includes the particular value, an additional iteration is performed.

32. The method of claim 31, wherein the controller is configured to operate the sensing circuitry to perform additional iterations by:

storing the remainder bit-vector as the zero dividend bit-vector;

comparing the remainder bit-vector with the second vector representing the number of divisors to determine whether the remainder bit-vector is greater than the second vector;

storing a result of the compare operation as the subtraction mask bit-vector;

subtracting the variable length elements of the second vector from the remainder bit-vector and store as the new remainder bit-vector using the subtraction mask bit-vector; and

incrementing the quotient bit-vector by a count of one corresponding to the subtraction mask bit-vector;

performing an XOR operation on the remainder bit-vector and the zero dividend bit-vector; and

store a result of the XOR operation in the number of corresponding compute components of the sensing circuitry.

33. The method of claim 32, wherein the method further includes repeating the additional iterations until the result stored in the number of corresponding compute components no longer stores the particular value.

34. A method for dividing variable length elements comprising:

performing a variable length division operation in a memory array on;

a plurality (M) of first elements stored in a first group of memory cells coupled to a first access line and to a number of sense lines of a memory array, wherein at least two of the plurality of first elements have different bit-lengths; and

a plurality (M) of second elements stored in the second group of memory cells coupled to a second access line and to the number of sense lines of the memory array, wherein at least two of the plurality of second elements have different bit-lengths; and

providing a division operation result that indicates a quotient of each corresponding element of the plurality of M first elements with each corresponding element of the plurality of M second elements.

35. The method of claim 34, wherein:

a first number of the M first elements and the M second elements are comprised of N bits and a second number of the M first elements and the M second elements are comprised of P bits; and

the first number of M first elements are divided by the corresponding first number of the M second elements and the second number of the M first elements are divided by the corresponding second number of the M second elements.

36. An apparatus, comprising:

a plurality of memory cells;

sensing circuitry comprising a plurality of sense amplifiers each coupled to a corresponding one of the memory cells and a plurality of compute components each coupled to a corresponding one of the sense amplifiers to implement a logic operation, and the sensing circuitry configured to temporarily store elements of at least a dividend bit-vector and a divisor bit-vector that are used to perform a division operation.

37. The apparatus of claim 36, further comprising:

a controller producing a control signal to perform the division operation that each of elements of a quotient bit-vector and a remainder bit-vector is produced in association with corresponding ones of the elements of the dividend bit-vector and the divisor bit-vector.

38. The apparatus of claim 36, wherein each of the compute components comprises:

a plurality of transistors configured to implement the logic operation to perform the division operation.

39. The apparatus of claim 36, wherein each of the elements of the dividend bit-vector is configured to be identical in bit length to each other.

40. The apparatus of claim 36, wherein at least one of the elements of the dividend bit-vector is configured to be different in bit length from a remaining one of the elements of the dividend bit-vector.