IPC분류정보
국가/구분 |
United States(US) Patent
등록
|
국제특허분류(IPC7판) |
|
출원번호 |
US-0151116
(2008-05-02)
|
등록번호 |
US-8250341
(2012-08-21)
|
발명자
/ 주소 |
- Schulz, Kenneth R
- Rapp, John W
- Jackson, Larry
- Jones, Mark
- Cherasaro, Troy
|
출원인 / 주소 |
- Lockheed Martin Corporation
|
대리인 / 주소 |
|
인용정보 |
피인용 횟수 :
23 인용 특허 :
112 |
초록
▼
A pipeline accelerator includes a bus and a plurality of pipeline units, each unit coupled to the bus and including at least one respective hardwired-pipeline circuit. By including a plurality of pipeline units in the pipeline accelerator, one can increase the accelerator's data-processing performan
A pipeline accelerator includes a bus and a plurality of pipeline units, each unit coupled to the bus and including at least one respective hardwired-pipeline circuit. By including a plurality of pipeline units in the pipeline accelerator, one can increase the accelerator's data-processing performance as compared to a single-pipeline-unit accelerator. Furthermore, by designing the pipeline units so that they communicate via a common bus, one can alter the number of pipeline units, and thus alter the configuration and functionality of the accelerator, by merely coupling or uncoupling pipeline units to or from the bus. This eliminates the need to design or redesign the pipeline-unit interfaces each time one alters one of the pipeline units or alters the number of pipeline units within the accelerator.
대표청구항
▼
1. A method, comprising: sending, via a communication bus, first data and a second data to first and second ones, respectively, of a plurality of pipeline units, each pipeline unit including a respective hardwired pipeline;processing the first data with the first pipeline unit synchronized to a firs
1. A method, comprising: sending, via a communication bus, first data and a second data to first and second ones, respectively, of a plurality of pipeline units, each pipeline unit including a respective hardwired pipeline;processing the first data with the first pipeline unit synchronized to a first clock signal;processing the second data with the second pipeline unit synchronized to a second clock signal and asynchronously relative to the first pipeline unit while the first pipeline unit is processing the first data, the second clock signal being unrelated to the first clock signal;generating a message that includes the first data and that identifies the first pipeline unit as a recipient of the message, wherein sending the first data to the first pipeline unit comprises determining from an address in a header of the message that the first pipeline unit is a recipient of the message;receiving the first data from the communication bus with a hardwired-pipeline circuit;processing the received first data in response to a synchronization signal to cause the first pipeline unit to process the data if the first pipeline unit is the recipient of the message; and loading the processed first data into a memory. 2. The method of claim 1 wherein sending the data comprises: sending the data to a router; andproviding the first data to the first pipeline unit with the router via a respective first branch of the communication bus. 3. The method of claim 1 wherein sending the first data comprises sending the first data to the first pipeline unit with a processor. 4. The method of claim 1 wherein sending the first data comprises sending the first data to the first pipeline unit with a third of the plurality of pipeline units. 5. The method of claim 1, further comprising driving the processed first data onto the communication bus with the first pipeline unit. 6. The method of claim 1 wherein processing the first data with the first pipeline unit comprises: receiving the first data from the communication bus with a hardwired-pipeline circuit,loading the first data into a memory with the hardwired-pipeline circuit,retrieving the first data from the memory with the hardwired-pipeline circuit, andprocessing the retrieved data with the hardwired-pipeline circuit. 7. The method of claim 1, further comprising: wherein processing the first data with the first pipeline unit comprises:retrieving the processed first data from the memory and driving the processed data onto the communication bus with the hardwired-pipeline circuit. 8. A pipeline accelerator, comprising: a communication bus;a plurality of clock signal lines operable to carry at least a first clock signal and a second clock signal, the first and second clock signals being unrelated to one another; anda plurality of pipeline units each comprising a respective hardwired-pipeline and each coupled to the communication bus, wherein first and second pipeline units receive first data and a second data over the communications bus,wherein the first pipeline unit is operable to receive a message including the first data and including an address identifying that the first pipeline unit is the recipient of the message,wherein the first pipeline unit is operable to process the first data synchronized to the first clock signal,wherein the second pipeline unit is operable to process the second data synchronized to a second clock signal and asynchronously relative to the first pipeline unit while the first pipeline unit is processing the first data, the second clock signal being unrelated to the first clock signal,wherein the first pipeline unit is operable to process the received first data in response to a synchronization signal to cause the first pipeline unit to process the data when the first pipeline unit is the recipient of the message; andwherein the first pipeline unit is operable to load the processed first data into a memory. 9. The pipeline accelerator of claim 8, wherein each of the pipeline units comprises: a respective memory coupled to the hardwired-pipeline circuit; andwherein the hardwired-pipeline circuit is operable to:receive data from the communication bus,load the data into the memory,retrieve the data from the memory,process the retrieved data, anddrive the processed data onto the communication bus. 10. The pipeline accelerator of claim 8, wherein each of the pipeline units comprises: a respective memory coupled to the hardwired-pipeline circuit; andwherein the hardwired-pipeline circuit is operable to: receive data from the communication bus,process the data,load the processed data into the memory,retrieve the processed data from the memory, andload the retrieved data onto the communication bus. 11. The pipeline accelerator of claim 8, further comprising: a pipeline bus; anda pipeline-bus interface coupled to the communication bus and to the pipeline bus. 12. The pipeline accelerator of claim 8 wherein the communication bus comprises a plurality of branches, a respective branch coupled to each pipeline unit and further comprising a router coupled to each of the branches. 13. The pipeline accelerator of claim 12, further comprising: a pipeline bus; anda pipeline-bus interface coupled to the router and to the pipeline bus. 14. The pipeline accelerator of claim 13 further comprising a secondary bus coupled to the router.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.