Simulation device and simulation method therefor

Patent metadata

Field	Value
Publication number	US-10162913-B2
Application number	US-201214233610-A
Country	US
Kind code	B2
Filing date	Jul 20, 2012
Priority date	Jul 22, 2011
Publication date	Dec 25, 2018
Grant date	Dec 25, 2018

Abstract

The present invention relates to a simulation method and device. According to the present invention, a simulation method using a plurality of blocks comprises: a dividing step of dividing a simulation into computation operations for performing unique operations on the blocks and communication operations for data exchanges between different blocks; a grouping step of performing a grouping between the interdependent computation and communication operations; and a simulation performing step of performing an operation included in each group using the blocks according to whether or not the level of interdependency between the computation and communication operations is resolved.

First claim

The invention claimed is:

1. A method for performing simulation by a simulation device using a plurality of blocks, the method comprising:
extracting communication operations from a simulation operation, each of the communication operations being for exchanging data between different blocks, each block simulating an electronic element;
sorting the communication operations based on dependency between the communication operations;
extracting computation operations for the plurality of blocks, at least one computation operation of the computation operations being for performing function of a block associated with the communication operations;
grouping at least one computation operation and at least one communication operation into groups from the computation operations and the communication operations for the plurality of blocks, the at least one computation operation and the at least one communication operation included in one group being dependent on each other; and
executing operations included in each group using the plurality of blocks based on dependency between the operations.

2. The method of claim 1 , wherein the executing of the operations comprises:
selecting a certain block;
selecting computation operations to be executed by the selected block from respective groups;
selecting computation operation independent from communication operation and closest to next communication operation among the selected computation operations; and
executing the selected computation operation.

3. The method of claim 2 , wherein the executing of the operations comprises executing, when an execution time of the communication operation arrives in executing the computation operation, the communication operation.

4. A simulation device for performing simulation using a plurality of blocks, the device comprising:
a memory configured to store at least one group constituting the simulation; and
at least one processor configured to control a plurality of blocks when performing the simulation,
wherein the at least one processor configured to:
extract communication operations from a simulation operation, each of the communication operations being for exchanging data between different blocks, each block simulating an electronic element,
sort the communication operations based on dependency between the communication operations,
extract computation operations for the plurality of blocks, at least one computation operation of the computation operations being for performing function of a block associated with the communication operations,
group at least one computation operation and at least one communication operation into groups from the computation operations and the communication operations for the plurality of blocks, the at least one computation operation and the at least one communication operation included in one group being dependent on each other, and
execute operations included in each group using the plurality of blocks based on dependency between the operations.

5. The device of claim 4 , wherein the at least one processor is further configured to:
select a certain block,
select computation operations to be executed by the selected block from respective groups,
select computation operation independent from communication operation and closest to next communication operation among the selected computation operations, and
execute the selected computation operation.

6. The device of claim 5 , wherein the at least one processor is further configured to execute, when an execution time of the communication operation arrives in executing the computation operation, the communication operation.

7. A method for performing simulation by a simulation device in a distributed system including at least two nodes connected to each other and having a plurality of blocks, the method comprising:
configuring a block at each node, the block simulating an electronic element;
receiving, if an operation request is transmitted from a first node, the requested operation at the block included in the first node;
pre-processing, at the block included in the first node, the requested operation before the second node processes the requested operation;
transferring, at the block included in the first node, the operation request to the second node;
receiving, at the block included in the first node, a processing result transmitted by the second node, the processing result being generated by the second node; and
adjusting a pre-processing result based on comparing the processing result with the pre-processing result to optimize the simulation of the simulation device.

8. The method of claim 7 , further comprising:
updating at the block, the pre-processing result by receiving a processing result transmitted by the second node.

9. The method of claim 8 , wherein the updating comprises updating difference between time taken for performing pre-processed service and the time taken for performing post-processed service.

10. The method of claim 8 , wherein the pre-processing comprises performing, when the block receives an operation request for a memory, a read or write function.

11. The method of claim 8 , wherein the pre-processing comprises by-passing, when the block receives a computation request for an active device, the operation request.

12. The method of claim 8 , wherein the pre-processing comprises performing, when the block receives an operation request for a passive device, a service according to behavior model of the passive device.

13. A simulation device for performing simulation in a distributed system, the device comprising:
at least one processor; and
at least two nodes each including a plurality blocks, each block simulating an electronic element,
wherein the at least one processor is configured to:
receive, if an operation request is transmitted from a first node, the requested operation at a block included in the first node,
pre-process, at the block included in the first node, the requested operation before the second node processes the requested operation,
transfer, at the block included in the first node, the operation request to the second node,
receive, at the block included in the first node, a processing result transmitted by the second node, the processing result being generated by the second node, and
adjust a pre-processing result based on comparing the processing result with the pre-processing result to optimize the simulation of the simulation device.

14. The device of claim 13 , wherein the at least one processor is further configured to transfer, after pre-processing, the operation request from the block to the second node and update the pre-processing result by receiving a processing result transmitted by the second node.

15. The device of claim 14 , wherein the at least one processor is further configured to update difference between time taken for performing pre-processed service and the time taken for performing post-processed service.

16. The device of claim 15 , wherein the at least one processor is further configured to perform, when the block receives an operation request for a memory, a read or write function.

17. The device of claim 15 , wherein the at least one processor is further configured to by-pass, when the block receives a computation request for an active device, the operation request.

18. The device of claim 15 , wherein the at least one processor is further configured to perform, when an operation request is received for a passive device, a service according to behavior model of the passive device.

Assignees

• Kim Kyoung Hoon
• Kim Joong Baik
• Lee Seung Wook
• Samsung Electronics Co Ltd

Inventors

• Kim Kyoung Hoon
• Kim Joong Baik
• Lee Seung Wook

Classifications

• G06F11/3495 Primary

  for systems · CPC title

• G06F9/5066

  Algorithms for mapping a plurality of inter-dependent sub-tasks onto a plurality of physical CPUs (mappping at compile time, see G06F8/451) · CPC title

• G06F11/3457

  Performance evaluation by simulation · CPC title

• G06F2209/5017

  Task decomposition · CPC title

• G06F11/3419

  by assessing time · CPC title