Method and apparatus for reducing power dissipation in finite field arithmetic circuits

• US 6,662,346 B1
• Filed: 02/08/2002
• Issued: 12/09/2003
• Est. Priority Date: 10/03/2001
• Status: Expired due to Fees

• Current Assignee: Marvell International Limited (Marvell Technology Group Limited)
• Original Assignee: Marvell International Limited (Marvell Technology Group Limited)
• Inventors: Feng, Weishi, Yu, Zhan
• Primary Examiner(s): Siek, Vuthe
• Assistant Examiner(s): LEVIN, NAUM B
• Application Number:
  US10/072,034
• Time in Patent Office: 669 Days
• Field of Search: 716/1, 716/2-4, 708/319, 714/769, 382/181
• US Class Current: 716/109
• CPC Class Codes: G06F 2119/06 Power analysis or power optimisation
• CPC Class Codes: G06F 30/30 Circuit design
• CPC Class Codes: G06F 7/724 Finite field arithmetic for error detection or correction in general H03M13/00, in computers G06F11/10

Abstract

A finite field arithmetic circuit with reduced power dissipation has first and second circuit inputs. A first circuit transition probability of the first circuit input is calculated by applying a random input to the first circuit input and a constant input to the second circuit input. A second circuit transition probability of the second circuit input is calculated by applying a constant input to the first circuit input and a random input to the second circuit input. One of the first and second circuit inputs having a lower circuit transition probability is selected. A first time-varying rate that a first input signal to the arithmetic circuit varies is compared with a second time-varying rate that a second input signal to the arithmetic circuit varies. The input signal having a higher time-varying rate is selected and coupled to the selected one of the first and second circuit inputs.

Claims

1. A method for reducing power dissipation of a finite field arithmetic circuit having first and second circuit inputs, comprising:

calculating a first circuit transition probability of said first circuit input by applying a random input to said first circuit input and a constant input to said second circuit input;

calculating a second circuit transition probability of said second circuit input by applying a constant input to said first circuit input and a random input to said second circuit input;

comparing said first circuit transition probability to said second circuit transition probability;

selecting one of said first and second circuit inputs having a lower circuit transition probability;

comparing a first time-varying rate that a first input signal to said arithmetic circuit varies with a second time-varying rate that a second input signal to said arithmetic circuit varies; and

selecting one of said first and second input signals having a higher time-varying rate.

2. The method of claim 1 further comprising coupling said selected one of said first and second input signals to said selected one of said first and second circuit inputs.

3. The method of claim 2 further comprising coupling the other of said first and second input signals to the other of said first and second circuit inputs.

4. The method of claim 2 wherein said arithmetic circuit is a finite field multiplier.

5. The method of claim 4 wherein said finite field multiplier is one of a semi-systolic multiplier and a Mastrovito multiplier.

6. The method of claim 2 wherein said arithmetic circuit is a finite field multiplier implemented in a finite field divider.

7. The method of claim 6 wherein said finite field divider is implemented as an inverter that is connected to said finite field multiplier.

8. The method of claim 7 wherein said finite field multiplier is one of a semi-systolic multiplier and a Mastrovito multiplier and wherein said inverter is implemented directly in combinatorial logic.

9. The method of claim 2 wherein said arithmetic circuit is a finite field multiplier implemented in a composite field multiplier.

10. The method of claim 2 wherein said arithmetic circuit is a finite field multiplier implemented in a composite field divider.

11. The method of claim 2 wherein said arithmetic circuit is implemented in an error correcting coding circuit.

12. The method of claim 11 wherein said error correction coding circuit is a Reed-Solomon error correction coding circuit.

13. The method of claim 1 wherein said circuit transition probability of said first circuit input is determined by computing node transition probabilities of circuit nodes associated with said first circuit input.

14. The method of claim 1 wherein said circuit transition probability of said second circuit input is determined by computing node transition probabilities of circuit nodes associated with said second circuit input.

15. A finite field arithmetic circuit comprising:

a first circuit input having a first circuit transition probability;

a second circuit input having a second circuit transition probability, wherein said second circuit transition probability is lower that said first circuit transition probability;

a first input that generates a first input signal to said first circuit input of said finite field arithmetic circuit that has a first time-varying rate; and

a second input that generates a second input signal to said second circuit input of said finite field arithmetic circuit, wherein said second input signal has a second time-varying rate that is higher than said first time-varying rate.

16. The finite field arithmetic circuit of claim 15 wherein said arithmetic circuit is a finite field multiplier.

17. The finite field arithmetic circuit of claim 16 wherein said finite field multiplier is one of a semi-systolic multiplier and a Mastrovito multiplier.

18. The finite field arithmetic circuit of claim 15 wherein said arithmetic circuit is a finite field multiplier implemented in a finite field divider.

19. The finite field arithmetic circuit of claim 18 wherein said finite field divider is implemented as an inverter that is connected to said finite field multiplier.

20. The finite field arithmetic circuit of claim 19 wherein said finite field multiplier is one of a semi-systolic multiplier and a Mastrovito multiplier and wherein said inverter is implemented directly in combinatorial logic.

21. The finite field arithmetic circuit of claim 15 wherein said arithmetic circuit is a finite field multiplier implemented in a composite field multiplier.

22. The finite field arithmetic circuit of claim 15 wherein said arithmetic circuit is a finite field multiplier implemented in a composite field divider.

23. The finite field arithmetic circuit of claim 15 wherein said arithmetic circuit is implemented in an error correcting coding circuit.

24. The finite field arithmetic circuit of claim 23 wherein said error correction coding circuit is a Reed-Solomon error correction coding circuit.

25. A method for reducing power dissipation of a finite field arithmetic circuit having first and second circuit inputs, comprising:

calculating a first circuit transition probability of said first circuit input by applying a random input to said first circuit input and a constant input to said second circuit input;

calculating a second circuit transition probability of said second circuit input by applying a constant input to said first circuit input and a random input to said second circuit input;

comparing said first circuit transition probability to said second circuit transition probability;

selecting one of said first and second circuit inputs having a higher circuit transition probability;

comparing a first time-varying rate that a first input signal to said arithmetic circuit varies with a second time-varying rate that a second input signal to said arithmetic circuit varies; and

selecting one of said first and second input signals having a lower time-varying rate.

26. The method of claim 25 further comprising the step of coupling said selected one of said first and second input signals to said selected one of said first and second circuit inputs.

27. The method of claim 26 further comprising coupling the other of said first and second input signals to the other of said first and second circuit inputs.

28. The method of claim 26 wherein said arithmetic circuit is a finite field multiplier.

29. The method of claim 28 wherein said finite field multiplier is one of a semi-systolic multiplier and a Mastrovito multiplier.

30. The method of claim 26 wherein said arithmetic circuit is a finite field multiplier implemented in a finite field divider.

31. The method of claim 30 wherein said finite field divider is implemented as an inverter that is connected to said finite field multiplier.

32. The method of claim 31 wherein said finite field multiplier is one of a semi-systolic multiplier and a Mastrovito multiplier and wherein said inverter is implemented directly in combinatorial logic.

33. The method of claim 26 wherein said arithmetic circuit is a finite field multiplier implemented in a composite field multiplier.

34. The method of claim 26 wherein said arithmetic circuit is a finite field multiplier implemented in a composite field divider.

35. The method of claim 26 wherein said arithmetic circuit is implemented in an error correcting coding circuit.

36. The method of claim 35 wherein said error correction coding circuit is a Reed-Solomon error correction coding circuit.

37. The method of claim 25 wherein said circuit transition probability of said first circuit input is determined by computing node transition probabilities of circuit nodes associated with said first circuit input.

38. The method of claim 25 wherein said circuit transition probability of said second circuit input is determined by computing node transition probabilities of circuit nodes associated with said second circuit input.