초록 ▼

According to an embodiment of the invention, a computing machine comprises a pipeline accelerator, a host processor coupled to the pipeline accelerator, and a redundant processor, a redundant pipeline unit, or both, coupled to the host processor and to the pipeline accelerator. The computing machine

According to an embodiment of the invention, a computing machine comprises a pipeline accelerator, a host processor coupled to the pipeline accelerator, and a redundant processor, a redundant pipeline unit, or both, coupled to the host processor and to the pipeline accelerator. The computing machine may also include a system-restore server and a system-restore bus that allow the machine to periodically save the machine states in case of a failure. Such a computing machine has a fault-tolerant scheme that is often more flexible than conventional schemes. For example, if the pipeline accelerator has more extra “space” than the host processor, then one can add to the computing machine one or more redundant pipeline units that can provide redundancy to both the pipeline and the host processor. Therefore, the computing machine can include redundancy for the host processor even though it has no redundant processing units.

대표청구항 ▼

What is claimed is: 1. A computing machine, comprising: a pipeline bus operable to carry data; a system-restore bus that is separate from the pipeline bus; a pipeline accelerator operable to drive pipeline data onto the pipeline bus and to periodically drive onto the system-restore bus accelerator

What is claimed is: 1. A computing machine, comprising: a pipeline bus operable to carry data; a system-restore bus that is separate from the pipeline bus; a pipeline accelerator operable to drive pipeline data onto the pipeline bus and to periodically drive onto the system-restore bus accelerator information representative of an operational data state of the pipeline accelerator; a host processor separate from the pipeline accelerator operable to drive host data onto the pipeline bus and to periodically drive onto the system-restore bus processor information representative of an operational data state of the host processor; and a redundant processor separate from the pipeline accelerator and separate from the host processor operable to receive one of the pipeline data and the host data from the pipeline bus and to receive one of the pipeline information and the processor information from the system-restore bus in response to a failure of a respective one of the host processor and the pipeline accelerator. 2. The computing machine of claim 1, further comprising: wherein the pipeline accelerator comprises a pipeline unit; and a redundant pipeline unit having a redundant programmable integrated circuit and coupled to the pipeline bus. 3. An electronic system, comprising: a computing machine, comprising, a pipeline bus, a system-restore bus that is separate from the pipeline bus, a pipeline accelerator coupled to the pipeline and system-restore busses, a host processor separate from the pipeline accelerator coupled to the pipeline and system-restore busses, and a redundant processor separate from the pipeline accelerator and separate from the host processor coupled to the pipeline and system-restore busses; and a system restorer coupled to the computing machine via the system-restore bus. 4. A computing machine, comprising: i. a pipeline accelerator including a pipeline unit having a programmable-logic integrated circuit; ii. a system-restore bus coupled to the pipeline accelerator and coupled to a system restorer, the system-restore bus separate from the pipeline unit; iii. an instruction-executing host processor coupled to the pipeline accelerator; and iv. a redundant pipeline unit separate from the pipeline accelerator and separate from the host processor coupled to the pipeline accelerator, to the system-restore bus and to the host processor and including a redundant programmable-logic integrated circuit. 5. A system, comprising: a computing machine, comprising, a pipeline bus, a system-restore bus that is separate from the pipeline bus, a pipeline accelerator having a pipeline unit and coupled to the pipeline and system-restore busses, a host processor separate from the pipeline accelerator coupled to the pipeline and system-restore busses, and a redundant pipeline unit separate from the pipeline accelerator and separate from the host processor coupled to the pipeline and system-restore busses and to the host processor and including a redundant programmable integrated circuit; and a system restorer coupled to the computing machine via the system-restore bus. 6. A computing machine, comprising: a pipeline bus; a system-restore bus that is separate from the pipeline bus; a pipeline accelerator coupled to the pipeline and system-restore busses; a host processor separate from the pipeline accelerator coupled to the pipeline and system-restore busses; and a recovery device separate from the pipeline accelerator and separate from the host processor coupled to the system-restore bus and periodically operable to: save first data representing a data state of the pipeline accelerator during a respective predetermined period, and save second data representing a data state of the host processor during the respective predetermined period. 7. The computing machine of claim 6 wherein after a failure of the pipeline accelerator, the recovery device is further operable to restore to the pipeline accelerator the most recent first data. 8. The computing machine of claim 6 wherein after a failure of the pipeline accelerator, the recovery device is further operable to restore to the host processor the most recent second data. 9. The computing machine of claim 6 wherein after a failure of the host processor, the recovery device is further operable to restore to the pipeline accelerator the most recent first data. 10. The computing machine of claim 6 wherein after a failure of the host processor, the recovery device is further operable to restore to the host processor the most recent second data. 11. The computing machine of claim 6 wherein: the pipeline bus is operable to carry data; and the recovery device is coupled to the pipeline accelerator and to the host processor via the pipeline bus. 12. A system, comprising: a computing machine, comprising, a pipeline bus; a system-restore bus that is separate from the pipeline bus; a pipeline accelerator coupled to the pipeline and system-restore busses, a host processor separate from the pipeline accelerator coupled to the pipeline and system-restore busses, and a recovery device separate from the pipeline accelerator and separate from the host processor coupled to the system-restore bus and operable to, periodically save first data representing a data state of the pipeline accelerator during a respective predetermined period, and periodically save second data representing a data state of the host processor during the respective predetermined period. 13. A method, comprising: processing data with a programmable logic integrated circuit of a pipeline accelerator and with a host processor separate from the pipeline accelerator and coupled to the pipeline accelerator; periodically retrieving an operational data state of the host processor via a system-restore bus that is separate from a data bus; providing the operational data state of the host processor to a redundant processor via the system-restore bus if the host processor fails, the redundant processor separate from the pipeline accelerator and separate from the host processor coupled to the pipeline accelerator; and processing the data with the pipeline accelerator and theredundant processor if the host processor fails. 14. A method, comprising: processing data with a programmable logic integrated circuit of a pipeline accelerator and with a host processor separate from the pipeline accelerator and coupled to the pipeline accelerator; periodically retrieving an operational data state of the programmable logic integrated circuit via a system-restore bus that is separate from a data bus; providing the operational data state of the programmable logic integrated circuit to a redundant programmable logic integrated circuit separate from the pipeline accelerator and separate from the host processor via the system-restore bus if the programmable logic integrated circuit fails, the redundant programmable logic integrated circuit coupled to the pipeline accelerator and to the processor; and processing the data with the redundant programmable logic integrated circuit if the programmable logic integrated circuit fails. 15. A method, comprising: periodically saving via a system-restore bus first data representing a data state of a pipeline accelerator during a respective predetermined period, the system-restore bus separate from a data bus over which the host processor and pipeline accelerator transfer processed data; and periodically saving via the system-restore bus second data representing a data state of a host processor during the respective predetermined period, the host processor being coupled to and separate from the pipeline accelerator. 16. The method of claim 15, further comprising restoring to the pipeline accelerator after a failure of the pipeline accelerator the most recent first data via the system-restore bus. 17. The method of claim 15, further comprising restoring to the host processor after a failure of the pipeline accelerator the most recent second data via the system-restore bus. 18. The method of claim 15, further comprising restoring to the pipeline accelerator after a failure of the host processor the most recent first data via the system-restore bus. 19. The method of claim 15, further comprising restoring to the host processor after a failure of the host processor the most recent second data via the system-restore bus.