Loop accelerator and data processing system having the same
IPC분류정보
국가/구분 |
United States(US) Patent
등록
|
국제특허분류(IPC7판) |
|
출원번호 |
UP-0514889
(2006-09-05)
|
등록번호 |
US-7590831
(2009-09-24)
|
우선권정보 |
KR-10-2006-0000703(2006-01-03) |
발명자
/ 주소 |
- Ryu, Soo jung
- Kim, Jeong wook
- Kim, Suk jin
- Kim, Hong Seok
- Kong, Jun jin
|
출원인 / 주소 |
- Samsung Electronics Co., Ltd.
|
대리인 / 주소 |
|
인용정보 |
피인용 횟수 :
0 인용 특허 :
5 |
초록
▼
Provided are a loop accelerator and a data processing system having the loop accelerator. The data processing system includes a loop accelerator which executes a loop part of a program, a processor core which processes a remaining part of the program except the loop part, and a central register file
Provided are a loop accelerator and a data processing system having the loop accelerator. The data processing system includes a loop accelerator which executes a loop part of a program, a processor core which processes a remaining part of the program except the loop part, and a central register file which transmits data between the processor core and the loop accelerator. The loop accelerator includes a plurality of processing elements (PEs) each of which performs an operation on each word to execute the program, a configuration memory which stores configuration bits indicating operations, states, etc. of the PEs, and a plurality of context memories, installed in a column or row direction of the PEs, which transmit the configuration bits along a direction toward which the PEs are arrayed. Thus, a connection structure between the configuration memory and the PEs can be simplified to easily modify a structure of the loop accelerator so as to extend the loop accelerator.
대표청구항
▼
What is claimed is: 1. A data processing system comprising a loop accelerator which processes a loop part of a program, a processor core which processes a remaining part of the program except the loop part, and a storage which transmits data between the processor core and the loop accelerator, wher
What is claimed is: 1. A data processing system comprising a loop accelerator which processes a loop part of a program, a processor core which processes a remaining part of the program except the loop part, and a storage which transmits data between the processor core and the loop accelerator, wherein the loop accelerator comprises: a plurality of processing elements (PEs), each of which performs an operation on each word to process the program; a configuration memory which stores configuration bits indicating configuration information of the PEs; and a plurality of context memories installed in one of column and row directions of the PEs, wherein the plurality of context memories transmits the configuration bits provided from the configuration memory along a direction toward which the PEs are arrayed, wherein each of the context memories comprises: a shift register which temporarily stores the configuration bits provided from the configuration memory; a counter which counts a number of configuration bits provided to the shift register; and a comparator which compares the number of configuration bits with a number of the PEs. 2. The data processing system of claim 1, wherein the shift register comprises a plurality of storage spaces which store the configuration bits. 3. The data processing system of claim 2, wherein the shift register synchronizes with a clock signal to transmit the configuration bits to PEs closest to the shift register. 4. The data processing system of claim 3, wherein if one configuration bit is transmitted per one clock cycle and the number of configuration bits is one less than the number of the PEs, the shift register stops transmitting the configuration bits. 5. The data processing system of claim 4, wherein the PEs comprise storage spaces to transmit the configuration bits. 6. The data processing system of claim 5, wherein the PEs neighboring one another are connected to wires in order to transmit the configuration bits. 7. The data processing system of claim 6, wherein the transmission of the configuration bits from the context memories to the PEs is performed during executions of previous routines performed by the PEs. 8. The data processing system of claim 7, wherein if the PEs are arrayed in a row direction, the context memories are arrayed along the row direction of the PEs. 9. The data processing system of claim 8, wherein the number of context memories is equal to the number of rows of PEs. 10. The data processing system of claim 9, wherein the configuration bits provided from the context memories are transmitted to the PEs along the row direction of the PEs. 11. The data processing system of claim 7, wherein if the PEs are arrayed in a column direction, the context memories are arrayed along the column direction of the PEs. 12. The data processing system of claim 11, wherein the number of context memories is equal to the number of columns of PEs. 13. The data processing system of claim 11, wherein the configuration bits provided from the context memories are transmitted to the PEs along the column direction of the PEs.
이 특허에 인용된 특허 (5)
-
Dawes Robert L. (Allen TX), Adaptive processing system having an array of individually configurable processing components.
-
Garey, Kenneth E., Method and apparatus of selecting one of a plurality of predetermined configurations using only necessary bus widths based on power consumption analysis for programmable logic device.
-
Schmit,Herman, Programmable pipeline fabric utilizing partially global configuration buses.
-
Huppenthal,Jon M.; Guzy,D. James, Reconfigurable processor module comprising hybrid stacked integrated circuit die elements.
-
Mirsky, Ethan; French, Robert; Eslick, Ian, Three level direct communication connections between neighboring multiple context processing elements.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.