Dynamic FIFO for simulation

• US 20050080610A1
• Filed: 10/10/2003
• Published: 04/14/2005
• Est. Priority Date: 10/10/2003
• Status: Active Grant

• Current Assignee: Primary Integration LLC (Bureau Veritas SA)
• Original Assignee: Synopsys Incorporated
• Inventors: Bongoni, Srinivas, Kroll, Andrea, Groetker, Thorsten Heiner, Toma, Horia
• Granted Patent: US 7,346,483 B2
• Time in Patent Office: Days
• US Class Current: 703/22
• CPC Class Codes: G06F 2117/08 HW-SW co-design, e.g. HW-SW partitioning
• CPC Class Codes: G06F 2205/063 Dynamically variable buffer size
• CPC Class Codes: G06F 30/33 Design verification, e.g. functional simulation or model checking
• CPC Class Codes: G06F 5/06 for changing the speed of data flow, i.e. speed regularising or timing, e.g. delay lines, FIFO buffers; over- or underrun control therefor G06F7/78 takes precedence

Abstract

To perform a simulation, a design can be divided into “blocks” described by models. To ensure that data is efficiently transferred from an source model to a destination model, a dynamic first-in first-out (FIFO) can be placed between these models. The initial size of the dynamic FIFO can be set to a relatively small value. To prevent deadlock, the size of the FIFO can be automatically increased in size by increments. In this manner, the memory resources of the FIFO can be tightly controlled. Advantageously, the size of the optimized dynamic FIFO can be used as the desired size of the FIFO implemented in silicon, thereby also ensuring efficient use of silicon resources.

Claims

1. A simulation environment with a time component, the simulation environment comprising:

a first unit that produces data;

a second unit for consuming the data; and

a dynamic first-in first-out buffer (FIFO) that interfaces between the first unit and the second unit, the dynamic FIFO capable of automatically increasing its size to minimize deadlock.

2. The simulation environment of claim 1, wherein the dynamic FIFO determines whether a wait period has expired before increasing its size.

3. The simulation environment of claim 2, wherein the dynamic FIFO is further capable of automatically increasing the wait period if its size is increased a predetermined number of times.

4. The simulation environment of claim 1, wherein the dynamic FIFO has a timing behavior substantially similar to that of the second unit.

5. The simulation environment of claim 1, wherein the dynamic FIFO has a different timing behavior than that of the second unit.

6. A tool used in a simulation environment, the simulation environment including a first unit that produces data and a second unit for consuming the data, the tool comprising:

a dynamic first-in first-out buffer (FIFO) that interfaces between the first unit and the second unit, the dynamic FIFO capable of automatically increasing its size to eliminate deadlock between the first and second units.

7. The tool of claim 6, wherein the dynamic FIFO determines whether a wait period has expired before increasing its size.

8. The tool of claim 7, wherein the dynamic FIFO is further capable of automatically increasing the wait period if its size is increased a predetermined number of times.

9. The tool of claim 6, wherein the dynamic FIFO has a timing behavior different than that of the second unit.

10. A method of decreasing simulation time for a design, the simulation including a first model and a second model, the first model representing a first block of the design that generates a plurality of tokens, the second model representing a second block of the design that consumes the plurality of tokens based on a behavior, the method comprising:

providing a dynamic memory for

receiving the tokens,

storing the tokens, and

providing the tokens in an order output by the first model when requested by the second model,

setting a wait period of the dynamic memory, wherein the wait period is defined by a predetermined number of cycles associated with the behavior;

decreasing the wait period each time the dynamic memory is full and the second model does not request a token; and

automatically increasing a size of the dynamic memory by an increment if the wait period is zero.

11. The method of claim 10, further including resetting the wait period after increasing the size of the dynamic memory.

12. The method of claim 10, wherein the dynamic memory includes a first-in first-out buffer (FIFO).

13. The method of claim 10, further including setting a maximum size of the dynamic memory.

14. The method of claim 10, further including increasing a counter value each time the size of the memory increases.

15. The method of claim 14, further including automatically increasing the predetermined number of time units defining the wait period if the counter value exceeds a limit.

16. The method of claim 15, further including setting a maximum wait period for the dynamic memory.

17. The method of claim 15, wherein automatically increasing the size of the dynamic memory and automatically increasing the predetermined number of time units defining the wait period are alternately performed.

18. A dynamic memory useable in a simulation environment, the dynamic memory providing an interface between a first model and a second model, the first model generating a plurality of tokens, the second model consuming the plurality of tokens based on a behavior, the dynamic memory device comprising:

means for handling the tokens including receiving the tokens, storing the tokens, and providing the tokens in an order output by the first model when requested by the second model,

means for setting a wait period of the dynamic memory device, wherein the wait period is defined by a predetermined number of time units associated with the behavior;

means for decreasing the wait period each time the dynamic memory is full and the second model does not request a token; and

means for increasing a size of the means for storing the tokens by an increment if the wait period is zero.

19. The dynamic memory of claim 18, further including means for resetting the wait period after increasing the size of the dynamic memory.

20. The dynamic memory of claim 18, wherein the means for handling the tokens is implementable with a first-in first-out buffer (FIFO).

21. The dynamic memory of claim 18, further including means for setting a maximum size of the dynamic memory.

22. The dynamic memory of claim 18, further including means for increasing a counter value each time the size of the dynamic memory increases.

23. The dynamic memory of claim 22, further including means for increasing the predetermined number of cycles defining the wait period if the counter value exceeds a limit.

24. The dynamic memory of claim 23, further including means for setting a maximum wait period for the dynamic memory.

25. The dynamic memory of claim 24, further including means for alternately increasing the size of the dynamic memory and increasing the predetermined number of cycles defining the wait period.