Calculation unit anomaly detection processing method and device of coherent light quantum computer

Abstract

The invention relates to a computing unit anomaly detection processing method and device of a coherent light quantum computer. The calculation unit anomaly detection processing method of the coherent light quantum computer comprises the following steps: acquiring output information of a calculation unit during working, wherein the output information comprises abnormal state identification information of the calculation unit; extracting abnormal state identification information from the output information; under the condition that the abnormal state identification information meets the preset abnormal condition, it is determined that the calculation unit works abnormally; and under the condition that the calculation unit works abnormally, the calculation unit is controlled to execute a re-initialization operation. According to the calculation unit anomaly detection processing method of the coherent light quantum computer, the technical problem that the reliability of a calculation result becomes poor due to the fact that the anomaly is difficult to find and process in time when the coherent light quantum computer is used in the prior art is solved.

Description

Technical Field

[0001] This invention relates to the field of optical quantum computing technology, and in particular to a method and device for detecting and processing anomalies in the computing units of a coherent optical quantum computer. Background Technology

[0002] The computing unit (such as FPGA) is the core control unit of quantum computing. It is responsible for operations such as timing generation, signal processing, and computation scheduling. If its state is abnormal, it will directly lead to the failure of quantum computing.

[0003] The coherent optical quantum computer control system in related technologies lacks an automatic detection and processing mechanism for computing unit anomalies, making it difficult to detect anomalies such as computing unit failures in a timely manner. Consequently, it is unable to process the anomalies immediately, resulting in a decrease in the reliability of the computing results of the coherent optical quantum computer.

[0004] There is currently no effective solution to the technical problem of coherent optical quantum computers, which suffer from reduced reliability of computation results due to the difficulty in timely detection and handling of anomalies during use. Summary of the Invention

[0005] The present invention provides a method and device for detecting and processing anomalies in the computing units of a coherent optical quantum computer, which at least solves the technical problem in the related art that the reliability of the computing results of coherent optical quantum computers is deteriorated because it is difficult to detect and process anomalies in a timely manner during use.

[0006] To address the aforementioned problems, according to one aspect of the present invention, a method for detecting and processing anomalies in the computing unit of a coherent optical quantum computer is provided, comprising: acquiring output information of the computing unit during operation, the output information including anomaly status identification information of the computing unit; extracting the anomaly status identification information from the output information; determining that the computing unit has experienced an operational anomaly if the anomaly status identification information meets preset anomaly conditions; and controlling the computing unit to perform a re-initialization operation if the computing unit has experienced an operational anomaly.

[0007] In some embodiments, the output information includes first information, the abnormal state identification information includes first identification information, the first identification information is contained in the first information, and the first identification information characterizes the electronic control block state of the computing unit.

[0008] In some embodiments, the first identification information is a string in the first information that is in a preset character position; the abnormal status identification information meets the preset abnormal conditions including: the string of the first identification information is different from the first preset string, and the first preset string is the string corresponding to the normal state of the electronic control block.

[0009] In some embodiments, the output information includes second information representing the calculation result of the computing unit, and the abnormal state identification information includes second identification information, which is contained within the second information, and the second identification information represents the locked state of the computing unit.

[0010] In some embodiments, the second identification information is the string in the target character position of the second information; the abnormal status identification information meets the preset abnormal conditions including: the string of the second identification information is the same as the second preset string, and the second preset string is the string corresponding to the time when the computing unit is locked.

[0011] In some embodiments, when a computing unit malfunctions, the step of controlling the computing unit to perform a re-initialization operation includes: controlling the computing unit to stop the current computing operation and timing scheduling operation; and controlling the computing unit to re-initialize.

[0012] In some embodiments, when a malfunction occurs in the computing unit, the step of controlling the computing unit to perform a re-initialization operation further includes: setting the operating state of the variable optical attenuator to a no-attenuation state before controlling the computing unit to re-initialize; and restoring the operating state of the variable optical attenuator to the original attenuation state after controlling the computing unit to re-initialize.

[0013] In some embodiments, after the step of controlling the computing unit to perform a re-initialization operation in the event of a malfunction in the computing unit, the computing unit anomaly detection and processing method of the coherent optical quantum computer further includes: controlling the computing unit to perform the computing operation before the anomaly, and repeatedly executing the steps of acquiring output information, extracting anomaly status identification information, determining whether the computing unit has malfunctioned, and controlling the computing unit to perform a re-initialization operation in the event of a malfunction in the computing unit.

[0014] In some embodiments, the method for detecting and processing anomalies in the computing unit of a coherent optical quantum computer includes: issuing a fault alert when a preset fault condition is met, wherein the preset fault condition includes at least one of the following: the number of times the re-initialization operation is repeatedly executed reaches a first preset number, the number of timeouts in the response feedback of the request information sent to the computing unit reaches a second preset number, and the waiting time when the computing unit performs a computing operation reaches a preset duration.

[0015] In some embodiments, the anomaly detection and processing method of the computing unit of a coherent optical quantum computer satisfies at least one of the following: the output information is obtained based on a publish-subscribe mechanism and / or a request-response mechanism; and the communication with the computing unit is implemented based on ZMQ.

[0016] In one aspect, an anomaly detection and processing device for a coherent optical quantum computer is provided, comprising: an acquisition module for acquiring output information of the computing unit during operation, the output information including anomaly status identifier information of the computing unit; an extraction module for extracting the anomaly status identifier information from the output information; a determination module for determining that the computing unit has experienced an operational anomaly if the anomaly status identifier information meets preset anomaly conditions; and a control module for controlling the computing unit to perform a re-initialization operation when the computing unit experiences an operational anomaly.

[0017] In one aspect of this invention, a non-transitory machine-readable medium storing computer instructions is provided, the computer instructions being used to cause a computer to execute the computational unit anomaly detection processing method of any of the coherent optical quantum computers described above.

[0018] In one aspect of this invention, a computer program product is provided, including a computer program / instruction, which, when executed by a processor, implements the computational unit anomaly detection and processing method of any of the coherent optical quantum computers described above.

[0019] According to another aspect of the present invention, a computing unit anomaly detection and processing device for a coherent optical quantum computer is provided, comprising: a processor, and a memory storing a program, the program including instructions, which, when executed by the processor, cause the processor to perform any of the above-described methods for computing unit anomaly detection and processing of a coherent optical quantum computer.

[0020] The method for detecting and processing anomalies in the computing units of a coherent optical quantum computer according to this invention includes: acquiring the output information of the computing unit during operation, the output information including anomaly status identifier information of the computing unit; extracting the anomaly status identifier information from the output information; determining that the computing unit has experienced an operational anomaly when the anomaly status identifier information meets preset anomaly conditions; and controlling the computing unit to perform a re-initialization operation when the computing unit has experienced an operational anomaly. This embodiment extracts anomaly status identifier information (information indicating whether the computing unit is abnormal) based on the output information of the computing unit during operation and determines whether the computing unit has experienced an anomaly based on the anomaly status identifier information. In the event of an anomaly, re-initialization is triggered, achieving automated identification and recovery of computing unit operational anomalies without real-time manual intervention, significantly shortening downtime caused by computing unit anomalies. Simultaneously, it can promptly detect computing unit anomalies and trigger recovery, effectively preventing OPO pulse phase-locking failure and measurement feedback loop breakage caused by computing unit anomalies, avoiding problems such as quantum computing solution failure, shortened coherence time, and prolonged convergence time, ensuring the computational continuity of the coherent optical quantum computer, and improving the overall solution success rate. The embodiments of the present invention solve the technical problem in the related art that the coherent optical quantum computer control system is difficult to detect and handle abnormalities in a timely manner due to the lack of an automatic detection and recovery mechanism for abnormal computing units, which is conducive to ensuring the reliable operation of the coherent optical quantum computer.

[0021] Details of one or more embodiments of the present invention are set forth in the following drawings and description, so that other features, objects and advantages of the invention will be more readily understood. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other embodiments can be obtained based on these drawings without creative effort.

[0023] Figure 1 This is a schematic flowchart of an embodiment of the anomaly detection and processing method for the computing unit of the coherent optical quantum computer of the present invention.

[0024] Figure 2 This is a flowchart illustrating a specific embodiment of the anomaly detection and processing method for the computing unit of the coherent optical quantum computer of the present invention.

[0025] Figure 3 This is a schematic diagram of the abnormal detection and processing method of the computing unit of the coherent optical quantum computer of the present invention, showing the abnormal judgment process and automatic recovery process (re-initialization process) of the computing unit.

[0026] Figure 4 This is a schematic diagram of the abnormal detection and processing process of the computing unit in an embodiment of the coherent optical quantum computer of the present invention. Detailed Implementation

[0027] The following will refer to... Figures 1 to 4 Embodiments of the invention will be described in more detail below. While some embodiments of the invention are shown in the accompanying drawings, it should be understood that the invention can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the invention. It should be understood that the accompanying drawings and embodiments are for illustrative purposes only and are not intended to limit the scope of protection of the invention.

[0028] To address the aforementioned problems, embodiments of the present invention provide a method for detecting and processing anomalies in the computing units of a coherent optical quantum computer, such as... Figure 1 As shown, the method mainly includes: Step S101: Obtain the output information of the computing unit when it is working. The output information includes the abnormal status identification information of the computing unit.

[0029] Step S102: Extract the abnormal status identification information from the output information.

[0030] Step S103: If the abnormal status identification information meets the preset abnormal conditions, it is determined that the computing unit has experienced a working abnormality.

[0031] Step S104: In the event of a malfunction in the computing unit, control the computing unit to perform a re-initialization operation.

[0032] The method for detecting and processing anomalies in the computing units of a coherent optical quantum computer according to this invention includes: acquiring the output information of the computing unit during operation, the output information including anomaly status identifier information of the computing unit; extracting the anomaly status identifier information from the output information; determining that the computing unit has experienced an operational anomaly when the anomaly status identifier information meets preset anomaly conditions; and controlling the computing unit to perform a re-initialization operation when the computing unit has experienced an operational anomaly. This embodiment extracts anomaly status identifier information (information indicating whether the computing unit is abnormal) based on the output information of the computing unit during operation and determines whether the computing unit has experienced an anomaly based on the anomaly status identifier information. In the event of an anomaly, re-initialization is triggered, achieving automated identification and recovery of computing unit operational anomalies without real-time manual intervention, significantly shortening downtime caused by computing unit anomalies. Simultaneously, it can promptly detect computing unit anomalies and trigger recovery, effectively preventing OPO pulse phase-locking failure and measurement feedback loop breakage caused by computing unit anomalies, avoiding problems such as quantum computing solution failure, shortened coherence time, and prolonged convergence time, ensuring the computational continuity of the coherent optical quantum computer, and improving the overall solution success rate. The embodiments of the present invention solve the technical problem in the related art that the coherent optical quantum computer control system is difficult to detect and handle abnormalities in a timely manner due to the lack of an automatic detection and recovery mechanism for abnormal computing units, which is conducive to ensuring the reliable operation of the coherent optical quantum computer.

[0033] In coherent optical quantum computing, the computing unit is the core control unit, responsible for performing operations such as timing generation, signal processing, and computation scheduling. Its specific form can vary, as long as it can achieve the above functions. For example, it can be an FPGA, a dedicated ASIC chip, a heterogeneous computing SoC, an industrial-grade high-end DSP, a quantum computing-specific control MCU / MPU, a programmable logic controller, and so on.

[0034] In actual implementation, the output information can be diverse, as long as it can contain abnormal state identification information that indicates whether the computing unit is abnormal.

[0035] For example, in some optional embodiments, the output information includes first information, and the abnormal state identification information includes first identification information. The first identification information is contained within the first information, and the first identification information characterizes the state of the electronic control block of the computing unit. In this embodiment, the output information includes the first information, and the first information includes the first identification information characterizing the state of the electronic control block. This can be used as a basis for judging whether the state of the electronic control block is abnormal, thereby enabling targeted identification of abnormalities at the underlying hardware control level of the computing unit, accurately locating the working state of the core control module (electronic control block) of the computing unit, and providing accurate and specific identification basis for early detection of abnormalities in the computing unit. This avoids the problem of failure of computing unit timing generation and signal processing functions due to the inaccurate identification of abnormal electronic control block states, and ensures the state stability of the computing unit from the bottom layer.

[0036] In some specific embodiments, the first identification information is a string located at a preset character position in the first information; the abnormal state identification information meets preset abnormal conditions including: the string of the first identification information is different from the first preset string, and the first preset string is the string corresponding to the normal state of the electronic control block. By using the string at the preset character position as the state identifier of the electronic control block of the computing unit, and judging the abnormality based on the comparison result of the string feature bit and the preset normal string, on the one hand, it realizes the millisecond-level rapid judgment of the electronic control block state, with a fast detection response speed, which can move the anomaly detection forward, significantly shorten the solution failure window of quantum computing, and adapt to the core requirement of coherent optical quantum computers for low latency anomaly detection; on the other hand, the string character position-based verification method is simple and efficient, without complex calculations, and the customized character position identifier can accurately match the firmware design logic of the computing unit, improving the accuracy of the electronic control block state anomaly judgment and reducing the probability of false judgment.

[0037] For example, in a specific embodiment, the computing unit is an FPGA. The first information is a custom status code published by the FPGA through fpga_interface_event, and the first identification information is the last 13 bits of the suffix, which represents the status of the control block. When all 13 bits are 1 (the first preset string), it means the FPGA is normal; otherwise, it means the FPGA is abnormal. Of course, this is only an example, and the specific format of the first information and the first identification information, the position of the characters, etc., can be flexibly selected according to the actual situation.

[0038] For example, in some optional embodiments, the output information includes second information representing the computation result of the computing unit, and the abnormal state identification information includes second identification information, which is contained within the second information. The second identification information represents the locked state of the computing unit. By extracting the second identification information from the second information representing the computation result, targeted detection of the locked state at the computation execution level of the computing unit is achieved. This enables early detection of locked-up faults in the computing unit, timely prevention of invalid solutions in the locked state, and avoidance of invalid computation continuously occupying core resources such as OPO pulse sequences and optical field modulation links, thereby improving the resource utilization efficiency of the coherent optical quantum computer. It should be noted that this detection is not mutually exclusive with the first identification information detection of the aforementioned electronic control block state; both can be used independently or simultaneously. When both are used simultaneously, a dual-channel anomaly identification mechanism can be formed, compensating for the shortcomings of single-state detection and reducing the probability of missed anomaly detection.

[0039] As a specific embodiment, the second identification information is the string at the target character position in the second information; the abnormal state identification information meets the preset abnormal conditions, including: the string of the second identification information is the same as the second preset string, and the second preset string is the string corresponding to the time when the computing unit is locked. Locking abnormalities are determined by matching the string at the target character position of the calculation result with the preset locking string. On the one hand, based on the inherent characteristic pattern of the calculation result, fast and accurate identification of the locking state is achieved. This method does not require complex numerical calculations and is suitable for the real-time detection requirements of quantum computing. On the other hand, the preset locking characteristic string is precisely matched with the custom abnormal code of the computing unit's solver firmware, improving the professionalism and accuracy of locking state judgment. It can promptly identify when the computing unit shows a locking trend, avoiding the continued execution of invalid calculations and reducing damage to the coherence of quantum computing.

[0040] For example, in a specific embodiment, the computing unit is an FPGA, and the second information is the result of calculating maxcut. The solution test can be triggered via FPGA.in.calculate_test, and the maxcut result can be obtained by listening to fpga_calculate_event. The second identifier is the first three bits of the maxcut result, defined as 124 or 125 (a second preset string), indicating that the FPGA is locked. Similarly, this is only an example; the specific format of the second information and the position of the characters can be flexibly selected according to the actual situation. Maxcut refers to solving the maximum cut problem, a classic problem in quantum computing (a type of Ising problem). In this embodiment, the result of the FPGA solving maxcut is used as the basis for anomaly detection, to determine whether the FPGA is functioning correctly through simulation testing.

[0041] In an optional embodiment, considering the strong coupling between the optical field parameters and the electronic control unit in coherent optical quantum computing, optical field disturbances (such as optical power fluctuations, phase shifts, and polarization state changes) can indirectly lead to abnormal output of the computing unit, which can easily be misjudged as a fault in the computing unit itself, causing invalid re-initialization, destroying the coherence of the optical field, and wasting quantum computing resources. At the same time, abnormalities in the computing unit itself can also affect the optical field control in reverse. Single-dimensional detection cannot distinguish between faults in the computing unit itself and false abnormalities caused by optical field disturbances. Therefore, the computing unit abnormality detection and processing method of the coherent optical quantum computer in this embodiment further includes: Acquire optical field parameter information, which includes parameters of the core optical devices for quantum computing linked to the computing unit (such as the optical power, phase, and repetition frequency of the OPO pulse, the attenuation value of the VOA, the detection light intensity of the balanced detector, and the optical path difference of the fiber delay line). A dual-dimensional feature fusion judgment model is constructed, which involves associating and mapping the abnormal state identification information (first identification information and / or second identification information) of the electrical signal dimension with the optical field parameter information of the optical field dimension, and pre-setting the optical field-electrical signal co-normal threshold range; In determining whether the abnormal state identification information meets the preset abnormal conditions, the following judgment logic is adopted: If only the electrical signal identification is abnormal and the optical field parameters are within the threshold range, it is determined that the preset abnormal conditions are met. In this case, it indicates that the computing unit itself is abnormal, and the subsequent re-initialization process is executed. If the electrical signal identification is abnormal and the optical field parameters exceed the threshold range, it is determined that the preset abnormal conditions are not met. In this case, it is a false abnormality caused by optical field disturbance. In this case, the coherent optical quantum computer's computing unit abnormality detection and processing method includes: first performing optical field parameter calibration (such as VOA optical power compensation, OPO phase-locked correction), and then re-detecting the computing unit state to avoid invalid re-initialization. If the optical field parameters are continuously abnormal and the calibration is ineffective, it is determined that the preset abnormal conditions are met. In this case, it is an optical field-electrical control coupling abnormality, triggering optical field-computing unit collaborative re-initialization.

[0042] By adopting the above processing method, abnormal linkage judgment between the light field and the electronic control unit in optical quantum computing can be realized, completely solving the misjudgment problem of single-dimensional detection, avoiding invalid re-initialization caused by light field disturbance, reducing the damage to the coherence of the light field, and ensuring the coherence time of quantum computing.

[0043] In this embodiment, when a computing unit malfunctions, the steps of controlling the computing unit to perform a re-initialization operation include: controlling the computing unit to stop the current computing operation and timing scheduling operation; and controlling the computing unit to re-initialize.

[0044] By first stopping the current computation and timing scheduling, and then performing the re-initialization stepwise operation, it is possible to effectively prevent the erroneous operation of the computing unit in an abnormal state from continuously damaging the OPO pulse sequence and measurement feedback closed loop, avoiding further failure of phase lock and further shortening of coherence time, and avoiding the risk of fault expansion from the process perspective. At the same time, the stepwise re-initialization operation has clear logic, which allows the computing unit to complete the internal state reset in an orderly manner, avoiding hardware configuration and register data disorder caused by direct re-initialization, and improving the effectiveness and orderliness of the re-initialization operation.

[0045] In some optional embodiments, in the event of a malfunction in the computing unit, the step of controlling the computing unit to perform a re-initialization operation further includes: setting the operating state of the variable optical attenuator to a non-attenuated state before controlling the computing unit to re-initialize; and restoring the operating state of the variable optical attenuator to its original attenuated state after controlling the computing unit to re-initialize. In other words, this embodiment performs a linked operation of setting the variable optical attenuator (e.g., VOA2) to non-attenuated and restoring its original attenuation before and after re-initialization. Specifically, setting the variable optical attenuator to a non-attenuated state before re-initialization ensures that the computing unit receives sufficient optical injection power during initialization, successfully completing self-calibration. This fundamentally solves the problem of recovery failure caused by insufficient initialization optical power, significantly improving the success rate of computing unit re-initialization (actual verification shows a success rate of over 95%). Restoring the original attenuated state after initialization allows the optical field modulation link to quickly return to the normal operating optical power conditions, ensuring the phase-locking accuracy and solution accuracy of subsequent quantum computing, and preventing abnormal optical power from affecting the operating parameters of the computing unit and the coherence of the coherent light.

[0046] In this embodiment, after the step of controlling the computing unit to perform a re-initialization operation when the computing unit malfunctions, the computing unit anomaly detection and processing method of the coherent optical quantum computer further includes: controlling the computing unit to perform the computing operation before the anomaly, and repeatedly executing the steps of acquiring output information, extracting anomaly status identification information, determining whether the computing unit has malfunctioned, and controlling the computing unit to perform a re-initialization operation when the computing unit malfunctions.

[0047] By re-initializing and re-computing the computing unit, and cyclically executing the entire anomaly detection process, a closed-loop processing logic of anomaly detection, automatic recovery, recovery verification, and continuous monitoring is formed. On one hand, this accurately verifies the recovery effect of the computing unit, ensuring that it truly returns to its normal operating state before formal quantum computing begins, avoiding secondary anomalies caused by incomplete recovery and improving the stability of quantum computing. On the other hand, the cyclic detection mechanism enables continuous real-time monitoring of the computing unit's operating state, promptly detecting potential secondary anomalies after recovery, further strengthening the state protection of the computing unit and adapting to the needs of long-term continuous computation in coherent optical quantum computers.

[0048] In this embodiment, the method for detecting and processing anomalies in the computing units of a coherent optical quantum computer includes: During the operation of the computing unit, breakpoint saving nodes are set according to the task stage and / or iteration step size of quantum computing. When an abnormal operation of the computing unit is detected, the breakpoint data is immediately saved in its entirety, including optical field parameters and power-off data. In one specific embodiment, this includes: the intermediate progress of quantum computing, the core configuration parameters of the computing unit, the key state parameters of the optical field link (e.g., OPO phase, optical power, VOA attenuation value, detector detection data), and characteristic information of the quantum state.

[0049] During the reinitialization of the computing unit, the basic operation of the optical field link is not stopped, the basic phase lock of the OPO pulse, the basic attenuation value of the VOA, and the optical path difference of the fiber delay line are maintained. Only the quantum computing tasks linked with the computing unit are suspended to avoid the loss of coherence caused by the complete interruption of the optical field link. After the computing unit is reinitialized and verified to be normal, it first performs a fast calibration of the light field state based on the saved light field parameters (restoring the light field state to the state at the breakpoint), and then continues to execute the quantum computing task from the breakpoint position where the anomaly occurred based on the saved computing breakpoint data, instead of recalculating from scratch.

[0050] With this approach, after a re-initialization operation, there is no need to re-execute the quantum computing task from scratch. This breakpoint saving and continuation mechanism significantly improves the efficiency of quantum computing, reduces the waste of light field resources and coherence time, and further improves the overall solution success rate of coherent optical quantum computers.

[0051] As some optional embodiments, the anomaly detection and handling method for the computing unit of a coherent optical quantum computer includes: issuing a fault alert when a preset fault condition is met. The preset fault condition includes at least one of the following: the number of times the re-initialization operation is repeatedly executed reaches a first preset number; the number of timeouts in the response feedback of the request information sent to the computing unit reaches a second preset number; and the waiting time for the computing unit to perform a computing operation reaches a preset duration. For example, the ZMQ request-response client is set to a default timeout of 5 seconds and 3 reconnection attempts; the default timeout for waiting for the computing test result is 30 seconds; the FPGA re-initialization process can be retried up to 5 times, and if it times out or fails, it enters the anomaly handling stage and issues a fault alert so that human intervention can be carried out in a timely manner.

[0052] By setting multiple preset fault conditions, such as the number of retries, response timeout, and calculation of waiting time, and issuing fault alerts when the conditions are met, the method achieves a fault-tolerant fallback design: First, it avoids invalid automatic retries from continuously occupying optical field resources and destroying optical field coherence, preventing the fault from escalating to irreversible damage to optical / hardware devices; second, it can promptly guide manual intervention to handle hardware physical faults, underlying firmware core errors, and other problems that automatic recovery cannot solve when automatic recovery is exhausted, taking into account both the automated operation requirements of coherent optical quantum computers and the safety of manual fallback, reducing unnecessary equipment downtime; third, the clearly defined fault judgment threshold makes the fault tolerance mechanism more standardized and accurately adapts to the high requirements of coherent optical quantum computers for hardware stability and optical field coherence.

[0053] In some other alternative embodiments, the anomaly detection processing method for the computing unit of a coherent optical quantum computer satisfies at least one of the following: output information is obtained based on a publish-subscribe mechanism and / or a request-response mechanism; communication with the computing unit is implemented based on ZMQ. Using a publish-subscribe mechanism and / or a request-response mechanism to obtain output information, or using ZMQ to implement communication with the computing unit, ensures the real-time, comprehensive, and stable nature of the computing unit's output information. The dual-channel mechanism of publish-subscribe and request-response enables both active querying of the computing unit's status information and passive real-time reception of computation results, ensuring comprehensive and timely data sources for anomaly detection. Furthermore, ZMQ, as a lightweight message communication framework, perfectly matches the low-latency communication requirements of coherent optical quantum computers with its high throughput and low latency characteristics, ensuring rapid transmission of anomaly detection commands and re-initialization control commands. Moreover, ZMQ's mature communication mode effectively avoids anomaly detection failure caused by communication interruptions, improving the overall operational stability of the method.

[0054] The following detailed description of the anomaly detection and processing method for the computing unit of the coherent optical quantum computer of the present invention is illustrated with a specific embodiment: This embodiment addresses the optical field manipulation scenario of a coherent Ising machine in a coherent optical quantum computer, constructing an automatic detection and recovery system for FPGA computing unit anomalies. This embodiment uses a time-multiplexed coherent Ising machine as the object, with an OPO center wavelength of 1550nm, a pump wavelength of 775nm, a pulse width of 10-50ps, a repetition frequency of 1GHz, and an injected optical power typically ranging from 10-50000μW based on power meter A readings. The coherent optical quantum computer includes an OPO pulse link, an optical fiber delay line, a balanced detector, a measurement feedback module, an FPGA control computing unit, a VOA2 variable optical attenuator, and a ZMQ communication management module. The system uses a dual-channel collaborative approach of ZMQ request-response and publish-subscribe to acquire FPGA status information and calculation results, and incorporates optical injection control (VOA2) into the initialization closed loop to ensure that the FPGA completes self-calibration when the optical field is fully injected.

[0055] This embodiment includes the following parts: Optical field modulation and measurement feedback link: The OPO pulse sequence is formed into time-multiplexed pulses through the fiber delay line. The measurement feedback result is obtained by balanced / zero difference detection and sent to the FPGA modulation and calculation unit for pulse modulation and feedback control of Ising solution.

[0056] Status string quick verification: The FPGA publishes a custom status code via fpga_interface_event. The system extracts the last 13 bits of the message.status field as the "ECU status" suffix. If the last 13 bits are all 1s, the FPGA is considered normal; otherwise, it is considered abnormal and enters the recovery process. This last 13-bit code is a custom health identifier in the FPGA firmware.

[0057] Anomaly detection in calculation results: The solver test is triggered via FPGA.in.calculate_test, and the maxcut result is obtained by listening to fpga_calculate_event. If the result prefix is ​​124 or 125, the FPGA is considered locked. This prefix is ​​a custom exception code in the FPGA solver firmware, used for early detection of lock-up.

[0058] Timeout and retry mechanism: ZMQ request-response client is set to a default timeout of 5s and 3 reconnection attempts; calculation test waiting results have a default timeout of 30s; FPGA reinitialization process can retry a maximum of 5 times, and if it times out or fails, it will enter exception handling and issue a fault reminder.

[0059] Automatic recovery and VOA2 optical power guarantee: After recovery is triggered, the current solution / timing schedule is paused first, and sys_init is executed to reinitialize and restore the working parameters (physical bit count, problem size, configuration registers). Before each sys_init, VOA2 is set to 0000 (attenuation is turned off, allowing injected light to pass through without attenuation) to ensure that the FPGA obtains sufficient optical power to complete the state calibration during initialization; after initialization, the original VOA2 attenuation value is restored, and calculation tests are performed again to confirm the recovery effect.

[0060] The specific steps are as follows: 1. Initialization and optical power guarantee: The system establishes a ZMQ request-response and publish-subscribe channel. Before executing FPGA.in.sys_init, VOA2 is set to 0000 (attenuation is turned off) to ensure that the injected optical power exceeds the FPGA detection threshold to complete the state calibration; after initialization, the original VOA2 attenuation value is restored.

[0061] 2. Quick Status Code Verification: Subscribe to fpga_interface_event and extract the last 13 digits from message.status as a custom health code. If all 13 digits are 1, it is considered normal; otherwise, it is considered abnormal and the recovery process begins.

[0062] 3. Calculation Result Deadlock Detection: Trigger FPGA.in.calculate_test to perform a solution test. Example parameters include matrix_filename as dens_1000_50%.csv, scaling_factor as 0.01, decimal_width as 6, iterations as 100, and threshold as 129670. If the prefix of the maxcut result is 124 or 125, a deadlock is determined.

[0063] 4. Anomaly Recovery Process: In case of an anomaly, pause the current solution timing, execute sys_init for reinitialization (maximum 5 retries, VOA2 is set to 0000 and restored to its original value before and after each retrieval), and restore the physical bit count and problem size configuration; after completion, restart the calculation test to confirm that the FPGA status is normal.

[0064] This embodiment directly applies to the measurement feedback and phase-locked link of the OPO pulse sequence. Anomaly detection and recovery target lock-up and state coding anomalies during the pulse solving process. Unlike the general health checks of traditional classical control systems, it better ensures coherence and solution success rate. This embodiment uses a 13-bit status string verification, extracting the last 13 bits of a custom health code from the FPGA event message and quickly comparing it with an all-1 template to achieve millisecond-level status judgment. Through 124 / 125 prefix lock-up identification, anomaly detection can be achieved based on the maxcut result mode, detecting FPGA lock-up in advance and triggering recovery, avoiding invalid solutions consuming OPO pulse resources. Through step-by-step automatic recovery, VOA2 optical power assurance, sys_init reinitialization, parameter recovery, and retesting are chained together to form a closed-loop recovery process, improving the recovery success rate.

[0065] Secondly, this invention also provides an anomaly detection and processing device for a coherent optical quantum computer's computing unit, comprising: an acquisition module for acquiring output information of the computing unit during operation, the output information including anomaly status identifier information of the computing unit; an extraction module for extracting the anomaly status identifier information from the output information; a determination module for determining that the computing unit has experienced an operational anomaly when the anomaly status identifier information meets preset anomaly conditions; and a control module for controlling the computing unit to perform a re-initialization operation when the computing unit experiences an operational anomaly.

[0066] In some embodiments, the output information includes first information, the abnormal state identification information includes first identification information, the first identification information is contained in the first information, and the first identification information characterizes the electronic control block state of the computing unit.

[0067] In some embodiments, the first identification information is a string in the first information that is in a preset character position; the abnormal status identification information meets the preset abnormal conditions including: the string of the first identification information is different from the first preset string, and the first preset string is the string corresponding to the normal state of the electronic control block.

[0068] In some embodiments, the output information includes second information representing the calculation result of the computing unit, and the abnormal state identification information includes second identification information, which is contained within the second information, and the second identification information represents the locked state of the computing unit.

[0069] In some embodiments, the second identification information is the string in the target character position of the second information; the abnormal status identification information meets the preset abnormal conditions including: the string of the second identification information is the same as the second preset string, and the second preset string is the string corresponding to the time when the computing unit is locked.

[0070] In some embodiments, the control module is used to: control the computing unit to stop the current computing operation and the timing scheduling operation; and control the computing unit to reinitialize.

[0071] In some embodiments, the control module is further configured to: set the operating state of the variable optical attenuator to a no-attenuation state before the control calculation unit is reinitialized; and restore the operating state of the variable optical attenuator to the original attenuation state after the control calculation unit is reinitialized.

[0072] In some embodiments, the computing unit anomaly detection and processing device of a coherent optical quantum computer further includes: after controlling the computing unit to perform a re-initialization operation when the computing unit malfunctions, controlling the computing unit to perform the computing operation before the anomaly, and repeatedly executing the steps of acquiring output information, extracting anomaly status identification information, determining whether the computing unit malfunctions, and controlling the computing unit to perform a re-initialization operation when the computing unit malfunctions.

[0073] In some embodiments, the computing unit anomaly detection and processing device of the coherent optical quantum computer further includes: a fault reminder module, used to issue a fault reminder when a preset fault condition is met, the preset fault condition including at least one of the following: the number of times the re-initialization operation is repeatedly executed reaches a first preset number, the number of timeouts in the response feedback of the request information sent to the computing unit reaches a second preset number, and the waiting time when the computing unit performs a computing operation reaches a preset time.

[0074] In some embodiments, the computational unit anomaly detection processing device of the coherent optical quantum computer satisfies at least one of the following: the output information is obtained based on a publish-subscribe mechanism and / or a request-response mechanism; and the communication with the computational unit is implemented based on ZMQ.

[0075] Since the aforementioned coherent optical quantum computer's computing unit anomaly detection and processing device corresponds to all the technical features of the aforementioned coherent optical quantum computer's computing unit anomaly detection and processing method, the aforementioned coherent optical quantum computer's computing unit anomaly detection and processing device possesses all the technical effects of the coherent optical quantum computer's computing unit anomaly detection and processing method. Since the technical effects of the coherent optical quantum computer's computing unit anomaly detection and processing method have been explained in detail above, they will not be repeated here.

[0076] In addition, embodiments of the present invention also provide a non-transient machine-readable medium storing computer instructions, which are used to cause the computer to execute the above-described coherent optical quantum computer's computational unit anomaly detection processing method.

[0077] Furthermore, this embodiment of the invention also provides a computer program product, including a computer program / instruction, which, when executed by a processor, implements the above-described method for detecting and processing anomalies in the computing unit of any coherent optical quantum computer.

[0078] Finally, this embodiment of the invention also provides a computing unit anomaly detection and processing device for a coherent optical quantum computer, comprising: a processor, and a memory storing a program, the program including instructions, which, when executed by the processor, cause the processor to perform any of the above-described computing unit anomaly detection and processing methods for a coherent optical quantum computer.

[0079] Computer programs for implementing the methods of embodiments of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0080] In the context of embodiments of the present invention, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable signal medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, or infrared systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0081] It should be noted that the term "comprising" and its variations used in the embodiments of the present invention are open-ended, meaning "including but not limited to". The term "based on" means "at least partially based on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". The modifications of "one" and "multiple" mentioned in the embodiments of the present invention are illustrative and not restrictive. Those skilled in the art should understand that, unless explicitly indicated otherwise in the context, they should be understood as "one or more".

[0082] The user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in the embodiments of the present invention are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of the relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation entry points are provided for users to choose to authorize or refuse.

[0083] The steps described in the method embodiments provided by this invention can be performed in different orders and / or in parallel. Furthermore, the method embodiments may include additional steps and / or omit the steps shown. The scope of protection of this invention is not limited in this respect.

[0084] The term "embodiment" in this specification refers to a specific feature, structure, or characteristic described in connection with an embodiment that may be included in at least one embodiment of the invention. The appearance of this phrase in various places in the specification does not necessarily imply the same embodiment, nor does it imply independence or alternativeity from other embodiments. The various embodiments in this specification are described in a related manner, with reference to each other for similar or identical parts. In particular, for apparatus, device, and system embodiments, since they are substantially similar to method embodiments, the description is relatively simple, and relevant details are referred to in the description of the method embodiments.

[0085] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of protection. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the appended claims.

Claims

1. A method for detecting and processing anomalies in the computing units of a coherent optical quantum computer, characterized in that, include: Obtain the output information of the computing unit during operation, the output information including the abnormal state identification information of the computing unit; Extract the abnormal status identification information from the output information; If the abnormal status identification information meets the preset abnormal conditions, it is determined that the computing unit has malfunctioned. In the event of a malfunction in the computing unit, the computing unit is controlled to perform a re-initialization operation.

2. The method for detecting and processing anomalies in the computing units of a coherent optical quantum computer according to claim 1, characterized in that, The output information includes first information, and the abnormal state identification information includes first identification information, which is contained within the first information. The first identification information characterizes the electronic control block state of the computing unit.

3. The method for detecting and processing anomalies in the computing units of a coherent optical quantum computer according to claim 2, characterized in that, The first identification information is the string in the first information that is located at a preset character position; The abnormal status identification information meets the preset abnormal conditions, including: the string of the first identification information is different from the first preset string, and the first preset string is the string corresponding to the normal state of the electronic control block.

4. The method for detecting and processing anomalies in the computing units of a coherent optical quantum computer according to claim 1, characterized in that, The output information includes second information characterizing the calculation result of the computing unit, and the abnormal state identification information includes second identification information, which is contained within the second information. The second identification information characterizes the locked state of the computing unit.

5. The method for detecting and processing anomalies in the computing units of a coherent optical quantum computer according to claim 4, characterized in that, The second identification information is the string located at the target character position in the second information; The abnormal status identification information meets the preset abnormal conditions, including: the string of the second identification information is the same as the second preset string, and the second preset string is the string corresponding to when the computing unit is locked.

6. The method for detecting and processing anomalies in the computing units of a coherent optical quantum computer according to claim 1, characterized in that, In the event of a malfunction in the computing unit, the steps for controlling the computing unit to perform a re-initialization operation include: Control the computing unit to stop the current computing operation and the timing scheduling operation; Control the reinitialization of the computing unit.

7. The method for detecting and processing anomalies in the computing units of a coherent optical quantum computer according to claim 6, characterized in that, In the event of a malfunction in the computing unit, the steps for controlling the computing unit to perform a re-initialization operation further include: Before re-initializing the computing unit, the operating state of the variable optical attenuator is set to a no-attenuation state; After the computing unit is reinitialized, the operating state of the variable optical attenuator is restored to the original attenuation state.

8. The method for detecting and processing anomalies in the computing units of a coherent optical quantum computer according to any one of claims 1 to 7, characterized in that, Following the step of controlling the computing unit to perform a re-initialization operation in the event of a malfunction in the computing unit, the computing unit malfunction detection and processing method of the coherent optical quantum computer further includes: The system controls the computing unit to perform the computational operations prior to the exception, and repeatedly executes the steps of obtaining the output information, extracting the exception status identification information, determining whether the computing unit has experienced a malfunction, and controlling the computing unit to perform a re-initialization operation if the computing unit experiences a malfunction.

9. The method for detecting and processing anomalies in the computing units of a coherent optical quantum computer according to claim 8, characterized in that, The anomaly detection and processing method for the computing unit of the coherent optical quantum computer includes: issuing a fault alert when a preset fault condition is met, wherein the preset fault condition includes at least one of the following: the number of times the re-initialization operation is repeatedly executed reaches a first preset number, the number of timeouts in the response feedback of the request information sent to the computing unit reaches a second preset number, and the waiting time when the computing unit performs a computing operation reaches a preset time.

10. The method for detecting and processing anomalies in the computing units of a coherent optical quantum computer according to any one of claims 1 to 7, characterized in that, The anomaly detection and processing method for the computing unit of the coherent optical quantum computer satisfies at least one of the following: The output information is obtained based on a publish-subscribe mechanism and / or a request-response mechanism; Communication with the computing unit is implemented based on ZMQ.

11. An anomaly detection and processing device for a coherent optical quantum computer's computing unit, characterized in that, include: The acquisition module is used to acquire the output information of the computing unit when it is working, and the output information includes the abnormal status identification information of the computing unit; The extraction module is used to extract the abnormal state identification information from the output information; The determination module is used to determine that the computing unit has malfunctioned when the abnormal status identification information meets the preset abnormal conditions. The control module is used to control the computing unit to perform a re-initialization operation in the event of a malfunction in the computing unit.

12. A non-transitory machine-readable medium storing computer instructions, characterized in that, The computer instructions are used to cause the computer to execute the computational unit anomaly detection and processing method of the coherent optical quantum computer according to any one of claims 1 to 10.

13. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instructions are executed by the processor, they implement the computational unit anomaly detection and processing method of the coherent optical quantum computer according to any one of claims 1 to 10.

14. An anomaly detection and processing device for a coherent optical quantum computer's computing unit, comprising: A processor and a memory, wherein the memory stores a program, characterized in that the program includes instructions that, when executed by the processor, cause the processor to perform the computational unit anomaly detection and processing method of a coherent optical quantum computer according to any one of claims 1 to 10.

Images