System, method, and computer program product for implementing large integer operations on a graphics processing unit

1. A method, comprising:

compiling, via a processor, a source code linked to a large integer library to generate an executable file; and

executing, via the processor, the executable file to perform a large integer operation using a parallel processing unit,

wherein the large integer library includes functions for processing large integers that are optimized for the parallel processing unit, and wherein the functions include at least one function for an addition operation that generates a large integer result by summing a first large integer operand and a second large integer operand, the addition operation performed, at least in part, by the parallel processing unit by:

dividing each of the first large integer operand and the second large integer operand into a plurality of portions, wherein a size of each portion in the plurality of portions is less than or equal to an integer pipeline width in the cores of the parallel processing unit;

for each corresponding pair of portions divided from the first large integer operand and the second large integer operand, summing the portion from the first large integer operand to the portion from the second large integer operand to generate a result associated with the corresponding pair of portions;

determining whether the result for each corresponding pair of portions generates a early bit; and

propagating the carry bits generated by summing each corresponding pair of portions in the result.

2. The method of claim 1 , wherein each of the large integers comprise at least 128 bits.

3. The method of claim 1 , wherein each of the large integers comprise greater than 1024 bits.

4. The method of claim 1 , wherein the large integer library is included as a header file linked to the source code.

5. The method of claim 1 , wherein the large integer library is a binary file that is dynamically linked to the source code.

6. The method of claim 1 , wherein the functions for processing large integers are optimized for the parallel processing unit based on a number of cores within the parallel processing unit.

7. The method of claim 6 , wherein the large integer operation is performed by:

loading the first large integer operand and the second large integer operand into a local memory associated with the parallel processing unit; and

executing a plurality of threads on at least two cores in parallel to generate the result for the large integer operation.

8. The method of claim 1 , wherein the large integer library comprises an addition operation, a subtraction operation, a multiplication operation, a swap operation, a logical OR operation, and a logical AND operation.

9. A non-transitory computer-readable storage medium storing instructions that, when executed by a processor, cause the processor to perform steps comprising:

compiling, via a processor, a source code linked to a large integer library to generate an executable file; and

executing, via the processor, the executable file to perform a large integer operation using a parallel processing unit,

wherein the large integer library includes functions for processing large integers that are optimized for the parallel processing unit, and wherein the functions include at least one function for an addition operation that generates a large integer result by summing a first large integer operand and a second large integer operand, the addition operation performed, at least in part, by the parallel processing unit by:

dividing each of the first large integer operand and the second large integer operand into a plurality of portions, wherein a size of each portion in the plurality of portions is less than or equal to an integer pipeline width in the cores of the parallel processing unit;

for each corresponding pair of portions divided from the first large integer operand and the second large integer operand, summing the portion from the first large integer operand to the portion from the second large integer operand to generate a result associated with the corresponding pair of portions;

determining whether the result for each corresponding pair of portions generates a carry bit; and

propagating the carry bits generated by summing each corresponding pair of portions in the result.

10. The computer-readable storage medium of claim 9 , wherein the large integer library is included as a header file linked to the source code.

11. The computer-readable storage medium of claim 9 , wherein the functions for processing large integers are optimized for the parallel processing unit based on a number of cores within the parallel processing unit.

12. The computer-readable storage medium of claim 11 , wherein the large integer operation is performed by:

loading the first large integer operand and the second large integer operand into a local memory associated with the parallel processing unit; and

executing a plurality of threads on at least two cores in parallel to generate the result for the large integer operation.

13. A system, comprising;

a memory storing a large integer library and a source code linked to the large integer library;

a processor configured to:

compile the source code linked to a large integer to generate an executable file, and

execute the executable file to perform a large integer operation utilizing a parallel processing unit; and

the parallel processing unit configured to perform large integer operations,

wherein the large integer library includes functions for processing large integers that are optimized for the parallel processing unit, and wherein the functions include at least one function for an addition operation that generates a large integer result by summing a first large integer operand and a second large integer operand, the addition operation performed, at least in part, by the parallel processing unit by:

dividing each of the first large integer operand and the second large integer operand into a plurality of portions, wherein a size of each portion in the plurality of portions is less than or equal to an integer pipeline width in the cores of the parallel processing unit;

for each corresponding pair of portions divided from the first large integer operand and the second large integer operand, summing the portion from the first large integer operand to the portion from the second large integer operand to generate a result associated with the corresponding pair of portions;

determining whether the result for each corresponding pair of portions generates a carry bit; and

propagating the carry bits generated by summing each corresponding pair of portions in the result.

14. The system of claim 13 , wherein the parallel processing unit includes a streaming multiprocessor having a shared mernory accessible by a plurality of cores.

15. The system of claim 14 , wherein the first large integer operand and the second large integer operand are stored in the shared memory.

16. The system of claim 13 , wherein the functions for processing large integers are optimized for the parallel processing unit based on a number of cores within the parallel processing unit.

17. The system of claim 13 , wherein the large integer operation is performed by:

loading the first large integer operand and the second large integer operand into a local memory associated with the parallel processing unit; and

executing a plurality of threads on at least two cores in parallel to generate the result for the large integer operation.