A system and method of rendering three-dimensional (3D) graphics. The system for rendering 3D graphics may include a plurality of cores including a scratch pad memory, a first memory to perform a control flow, a second memory for loop acceleration, and a shared memory to interpolate with the plurali
A system and method of rendering three-dimensional (3D) graphics. The system for rendering 3D graphics may include a plurality of cores including a scratch pad memory, a first memory to perform a control flow, a second memory for loop acceleration, and a shared memory to interpolate with the plurality of cores.
대표청구항▼
1. A system for rendering three-dimensional (3D) graphics, the system comprising: a plurality of cores including a scratch pad memory and each selectively performing one of a control flow and loop acceleration based on functional characteristics of each rendering process;a first memory to perform a
1. A system for rendering three-dimensional (3D) graphics, the system comprising: a plurality of cores including a scratch pad memory and each selectively performing one of a control flow and loop acceleration based on functional characteristics of each rendering process;a first memory to perform a control flow;a second memory for loop acceleration; anda shared memory to interoperate with the plurality of cores,wherein data required for calculation of each of the plurality of cores is pre-fetched and stored in the shared memory. 2. The system of claim 1, wherein each of the plurality of cores is configured as a dual core as a basic unit and includes at least one dual core. 3. The system of claim 1, wherein each of the plurality of cores includes a reconfigurable processor core, and the first memory includes an instruction memory performing the control flow, and the second memory includes a configuration memory for the loop acceleration. 4. The system of claim 3, wherein the reconfigurable processor core includes a plurality Arithmetic Logic Units (ALUs), and all or a portion of the plurality of ALUs is operated depending on at least one predetermined mode. 5. The system of claim 4, wherein the at least one predetermined mode comprises: a Very Long Instruction Word (VLIW) mode to perform the control flow; anda Coarse Grained Array (CGA) mode for the loop acceleration,wherein the shared memory and the scratch pad memory are used in the VLIW mode, and the scratch pad memory is used in the CGA mode. 6. The system of claim 1, wherein each of the first memory and the second memory corresponds to each of the plurality of cores. 7. A method of tile-based rendering of 3D graphics, the method comprising: distributing data into a plurality of cores to selectively perform one of a control flow and loop acceleration; andprocessing the distributed data using any one of a plurality of predetermined modes based on functional characteristics of each rendering process,wherein data required for calculation of each of the plurality of cores is pre-fetched and stored in a shared memory. 8. The method of claim 7, wherein the distributing distributes any one of vertex data and triangle data into the plurality of cores. 9. The method of claim 7, wherein the plurality of cores includes at least one dual core, and the distributing divides a triangle into a top portion and a bottom portion and distributes the data into the at least one dual core. 10. The method of claim 7, wherein the plurality of modes comprises: a Very Long Instruction Word (VLIW) mode to perform control flow; anda Coarse Grained Array (CGA) mode for loop acceleration. 11. The method of claim 10, wherein the processing performs at least one of vertex shading, span processing, and texture mapping in the CGA mode, and performs at least one of sorting, testing, interpolation, determinant calculation, and division in the VLIW mode. 12. The method of claim 10, wherein the shared memory and a scratch pad memory are used in the VLIW mode, and the scratch pad memory is used in the CGA mode. 13. The method of claim 10, wherein a portion of ALUs of each of the plurality of cores is used in the VLIW mode, and all the ALUs of each of the plurality of cores is used in the CGA mode. 14. The method of claim 7, wherein the processing comprises: double-buffering data of a tile-based frame buffer using the shared memory in interoperation with the plurality of cores; andsequentially processing the data of the tile-based frame buffer. 15. At least one computer readable recording medium controlling at least one processing device to implement a method of tile-based rendering of 3D graphics, the method comprising: distributing data into a plurality of cores to selectively perform one of a control flow and loop acceleration; andprocessing the distributed data using any one of a plurality of predetermined modes based on functional characteristics of each rendering process,wherein data required for calculation of each of the plurality of cores is pre-fetched and stored in a shared memory.
연구과제 타임라인
LOADING...
LOADING...
LOADING...
LOADING...
LOADING...
이 특허에 인용된 특허 (10)
Diard,Franck R., Adaptive load balancing in a multi-processor graphics processing system.
Cooke Laurence H. ; Phillips Christopher E. ; Wong Dale, Method for compiling high level programming languages into an integrated processor with reconfigurable logic.
Musoll, Enrique; Nemirovsky, Mario; Huynh, Jeffrey, Packet processor that generates packet-start offsets to immediately store incoming streamed packets using parallel, staggered round-robin arbitration to interleaved banks of memory.
Putnam Leland K. (Salt Lake City UT) Lucht Phillip H. (Salt Lake City UT) Blair David K. (Salt Lake City UT) Pritchett Scott K. (Sandy UT), System and method for high speed computer graphics image computation using a parallel connected, asynchronous multiproce.
Pfister, Hanspeter; Kreeger, Kevin A.; Marks, Joseph W.; Shen, Chia, Temporal and spatial coherent ray tracing for rendering scenes with sampled and geometry data.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.