Computer structure pipeline презентация
Содержание
- 2. A Basic Processor
- 3. Pipelined Car Assembly
- 5. Pipelining Pipelining does not reduce the latency of single task, it
- 6. Pipelined CPU
- 7. Structural Hazard Different instructions using the same resource at the same
- 8. Pipeline Example: cycle 1
- 9. Pipeline Example: cycle 2
- 10. Pipeline Example: cycle 3
- 11. Pipeline Example: cycle 4
- 12. Pipeline Example: cycle 5
- 13. RAW Dependency
- 14. Using Bypass to Solve RAW Dependency
- 15. RAW Dependency
- 16. Forwarding Hardware
- 17. Forwarding Control Forwarding from EXE (L3) if (L3.RegWrite and (L3.dst ==
- 18. Register File Split Register file is written during first half of
- 19. Can't Always Forward
- 20. Stall If Cannot Forward
- 21. Software Scheduling to Avoid Load Hazards Fast code LW Rb,b LW
- 22. Control Hazards
- 23. Control Hazard on Branches
- 24. Control Hazard on Branches
- 25. Control Hazard on Branches
- 26. Control Hazard on Branches
- 27. Control Hazard on Branches
- 28. Control Hazard on Branches
- 29. Control Hazard: Stall Stall pipe when branch is encountered until resolved
- 30. Control Hazard: Predict Not Taken Execute instructions from the fall-through (not-taken)
- 31. Dynamic Branch Prediction
- 32. BTB Allocation Allocate instructions identified as branches (after decode) Both conditional
- 33. BTB (cont.) Wrong prediction Predict not-taken, actual taken Predict taken, actual
- 34. Adding a BTB to the Pipeline
- 35. Adding a BTB to the Pipeline
- 36. Adding a BTB to the Pipeline
- 37. Using The BTB
- 38. Using The BTB (cont.)
- 39. Backup
- 40. MIPS Instruction Formats
- 41. The Memory Space Each memory location is 8 bit =
- 42. Register File The Register File holds 32 registers Each register is
- 43. Memory Components Inputs Address: address of the memory location we wish
- 44. The Program Counter (PC) Holds the address (in memory) of the
- 45. Instruction Execution Stages Fetch Fetch instruction pointed by PC from I-Cache
- 46. The MIPS CPU
- 47. Executing an Add Instruction
- 48. Executing a Load Instruction
- 49. Executing a Store Instruction
- 50. Executing a BEQ Instruction
- 51. Control Signals
- 52. Pipelined CPU: Load (cycle 1 – Fetch)
- 53. Pipelined CPU: Load (cycle 2 – Dec)
- 54. Pipelined CPU: Load (cycle 3 – Exe)
- 55. Pipelined CPU: Load (cycle 4 – Mem)
- 56. Pipelined CPU: Load (cycle 5 – WB)
- 57. Datapath with Control
- 58. Multi-Cycle Control
- 59. Five Execution Steps Instruction Fetch Use PC to get instruction and
- 60. Five Execution Steps (cont.) Execution ALU is performing one of
- 61. The Store Instruction
- 62. RAW Hazard: SW Solution
- 63. Delayed Branch Define branch to take place AFTER n following instruction
- 64. Delayed Branch Performance Filling 1 delay slot is easy, 2 is
- 65. Скачать презентацию
Слайды и текст этой презентации