Automatic generation of streaming data interface circuit
원문보기
IPC분류정보
국가/구분
United States(US) Patent
등록
국제특허분류(IPC7판)
G06F-013/00
G06F-013/12
G06F-015/00
출원번호
UP-0231171
(2005-09-20)
등록번호
US-7603492
(2009-10-28)
발명자
/ 주소
Chai, Sek M.
Bellas, Nikos
Dwyer, Malcolm R.
Lau, Erica M.
Li, Zhiyuan
Linzmeier, Daniel A.
출원인 / 주소
Motorola, Inc.
인용정보
피인용 횟수 :
2인용 특허 :
31
초록▼
A streaming data interface device (700) of a streaming processing system (200) is automatically generated by selecting a set of circuit parameters (610) consistent with a set of circuit constraints and generating (612, 614) a representation of a candidate memory interface device based upon a set of
A streaming data interface device (700) of a streaming processing system (200) is automatically generated by selecting a set of circuit parameters (610) consistent with a set of circuit constraints and generating (612, 614) a representation of a candidate memory interface device based upon a set of stream descriptors. The candidate streaming data interface device is evaluated (616) with respect to one or more quality metrics and the representation of the candidate streaming processor circuit is output (622) if the candidate memory interface device satisfies a set of processing system constraints and is better in at least one of the one or more quality metrics than other candidate memory interface devices.
대표청구항▼
The invention claimed is: 1. A method for automatic design of a streaming data interface circuit of a processing system, the method comprising: selecting a set of circuit parameters consistent with a set of circuit constraints; obtaining a set of stream descriptors that describe data locations of t
The invention claimed is: 1. A method for automatic design of a streaming data interface circuit of a processing system, the method comprising: selecting a set of circuit parameters consistent with a set of circuit constraints; obtaining a set of stream descriptors that describe data locations of target data embedded within a data stream, wherein the target data can be transferred by the streaming data interface circuit between the data stream and a streaming device; generating a hardware description of a candidate streaming data interface circuit implemented as one of an application specific integrated circuit and a configurable logic, based upon the set of circuit parameters, that can transfer the target data as specified by the set of stream descriptors; evaluating the candidate streaming data interface circuit with respect to one or more quality metrics; and outputting the hardware description of the candidate streaming data interface circuit in response to the candidate streaming data interface circuit satisfying a set of system constraints and being better in at least one of the one or more quality metrics than other candidate streaming data interface circuits. 2. The method according to claim 1, wherein the obtaining of the set of stream descriptors comprises obtaining one or both of a set of source stream descriptors and destination stream descriptors, and wherein the target data can be transferred, respectively, from a source stream to the streaming circuit or from the streaming circuit to a destination stream, or both. 3. The method in accordance with claim 1, wherein the set of circuit constraints includes a constraint selected from the group consisting of the maximum area, the maximum number of buffer elements, the maximum circuit latency, and the maximum power dissipation for the streaming data interface circuit. 4. The method in accordance with claim 1, wherein the set of system constraints includes a constraint selected from the group consisting of bus width, bus latency, bus bandwidth, streaming circuit latency, and streaming circuit bandwidth. 5. A method in accordance with claim 1, wherein the steps of selecting a set of circuit parameters, obtaining a set of stream descriptors, generating a hardware description of a candidate streaming data interface circuit, and evaluating the candidate streaming data interface circuit are repeated to generate a set of hardware descriptions of pareto-optimal streaming data interface circuits that satisfy the set of system constraints and are better in at least one of the one or more quality metrics than other candidate streaming data interface circuits. 6. A method in accordance with claim 1, wherein a quality metric of the one or more quality metrics is a metric selected from the group consisting of the area of the candidate streaming data interface circuit, the latency of the candidate streaming data interface circuit, the throughput of the candidate streaming data interface circuit and the clock speed of the candidate streaming data interface circuit. 7. The method according to claim 1, wherein the set of stream descriptors includes a starting address and at least two of a STRIDE value, a SPAN value, a SKIP value, and a TYPE value. 8. The method in accordance with claim 1, wherein the hardware description of the candidate streaming processor circuit is a register transfer language hardware description. 9. The method in accordance with claim 1, wherein the stream data interface circuit is coupled in a system to at least one system component. 10. A streaming data interface circuit, comprising: a control registers unit to store stream descriptors and control parameters for at least one stream unit: at least one stream unit operable to generate addresses based on the stream descriptors and transfer target data between at least two system components as contiguous usable data using a data bus, wherein each of the at least one stream data units comprises a bus address queue to store the generated bus addresses for the target data; an arbiter unit operable to select the at least one stream unit for access to the data bus, wherein the selection is based on the control parameters, wherein the control registers unit, the at least one stream unit, and the arbiter unit are implemented as one of an application specific integrated circuit and a configurable logic. 11. The streaming data interface circuit according to claim 10, wherein each of the at least one stream unit comprises: an address generation unit to generate bus addresses for the target data and stream addresses for a system component based on stream descriptors; a line buffer unit to store the target data transferred on the data bus; a stream address queue to store the stream addresses for streaming data that is transferred between the stream unit and the system component; a stream data queue to store the streaming data; a ready unit to receive commands for transfer of the streaming data; and a select unit to transfer the proper amount of streaming data from the stream data queue according to the commands. 12. The streaming data interface circuit according to claim 10, wherein the set of stream descriptors includes a starting address and at least two of a STRIDE value, a SPAN value, a SKIP value, and a TYPE value. 13. The streaming data interface circuit in accordance with claim 10, wherein a direction of data transfer of the stream unit is configurable. 14. The streaming data interface circuit in accordance with claim 11, wherein a circuit in the address generation unit used to calculate bus addresses for the target data is configurable. 15. The streaming data interface circuit in accordance with claim 11, wherein a circuit in the address generation unit used to calculate stream addresses for a system component is configurable. 16. The streaming data interface circuit in accordance with claim 11, wherein a number of buffer elements of the stream address queue is configurable. 17. The streaming data interface circuit in accordance with claim 11, wherein a size of the stream address queue and a size of the stream data queue are configurable. 18. The streaming data interface circuit in accordance with claim 11, wherein a number of buffer elements of the line buffers unit is configurable. 19. The streaming data interface circuit in accordance with claim 11, wherein a width of ready unit and select unit to set the number of stream elements transferred per cycle is configurable. 20. A computer readable medium for storing a set of program instructions, which when executed on a computer, designs a streaming data interface circuit of a system, the program instructions comprising: a set of instructions for inputting a set of circuit constraints; a set of instructions for inputting a set of system constraints; a set of instructions for inputting a set of stream descriptors that identify locations of target data; a set of instructions for selecting a set of circuit parameters consistent with the set of circuit constraints; a set of instructions for generating a hardware description of a candidate streaming data interface circuit implemented as one of an application specific integrated circuit and a configurable logic and optimized based upon the set of circuit parameters to transfer target data specified by the stream descriptors; a set of instructions for evaluating the candidate streaming data interface circuit with reference to one or more quality metrics; and a set of instructions for outputting the hardware description of the candidate streaming data interface circuit when the candidate streaming data interface circuit satisfies the set of system constraints and is better in at least one of the one or more quality metrics than other candidate streaming data interface circuits.
연구과제 타임라인
LOADING...
LOADING...
LOADING...
LOADING...
LOADING...
이 특허에 인용된 특허 (31)
Diard,Franck R., Adaptive load balancing in a multi-processor graphics processing system.
Michael S. Schlansker ; Vinod K. Kathail ; Greg Snider ; Shail Aditya Gupta ; Scott A. Mahlke ; Santosh Abraham, Automated design of processor systems using feedback from internal measurements of candidate systems.
Srinivasan,Sankaranarayanan; Leavesley, III,W. Story; McHugh,George L.; Wieland,Douglas P.; Kalman, III,Sandor S., Method and apparatus for implementing a pre-implemented circuit design for a programmable logic device.
Pascoe Robert A. (Grapevine TX) Ehnebuske David L. (Irving TX), Method of creating and detecting device independent controls in a presentation data stream.
Van Praet Johan Roland,BEX ; Lanneer Dirk,BEX ; Geurts Werner Gustaaf Theresia,BEX ; Goossens Gert Lodewijk Huibrecht,BEX, Method of generating code for programmable processors, code generator and application thereof.
Wilson, Robert P.; Maydan, Dror E.; Wang, Albert Ren-Rui; Lichtenstein, Walter D.; Tjiang, Weng Kiang, System and method for dynamically designing and evaluating configurable processor instructions.
De Oliveira Kastrup Pereira, Bernardo; Bink, Adrianus J.; Hoogerbrugge, Jan, System for executing computer program using a configurable functional unit, included in a processor, for executing configurable instructions having an effect that are redefined at run-time.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.