Optical non-linear arithmetic device, and test system and test method for optical non-linear arithmetic device

Abstract

The invention relates to an optical nonlinear arithmetic device and a testing system and method of the optical nonlinear arithmetic device, the optical nonlinear arithmetic device comprises a tunable laser and a nonlinear arithmetic unit, the nonlinear arithmetic unit works in a light amplification mode, and the tunable laser and the nonlinear arithmetic unit are connected through an optical fiber. The tunable laser is used for emitting optical signals with adjustable wavelength and power, the nonlinear arithmetic unit is used for carrying out nonlinear transformation on the optical signals and outputting nonlinear optical signals, and the nonlinear optical signals and the optical signals are in a nonlinear activation function relation. The optical nonlinear arithmetic device has the advantages of high speed and low nonlinear optical power threshold.

Description

Technical Field

[0001] This application relates to the field of optical neural network technology, and in particular to an optical nonlinear computing device, a testing system for the optical nonlinear computing device, and a testing method for the testing. Background Technology

[0002] With the rapid development of artificial intelligence and high-performance computing technologies, unprecedented demands have been placed on the energy efficiency and speed of computing systems. Against this backdrop, optical neural networks, due to their inherent advantages of high throughput, low latency, and low power consumption, have become an important research direction for breaking through the bottlenecks of traditional electronic computing. In neural networks, nonlinear activation functions are crucial for improving the network's depth and expressive power; therefore, achieving efficient and flexible nonlinear computation in the optical domain has become a core technical problem that urgently needs to be solved.

[0003] However, the nonlinear activation function obtained using the spatial light nonlinear scheme described in related technologies has a low matching degree with the nonlinear activation function in the neural network. Summary of the Invention

[0004] Therefore, it is necessary to provide an optical nonlinear computing device, a testing system for the optical nonlinear computing device, and a testing method to address the above-mentioned technical problems. The nonlinear activation function output by the optical nonlinear computing device has a high degree of matching with the nonlinear activation function in the network.

[0005] In a first aspect, this application provides an optical nonlinear computing device, which includes: a tunable laser and a nonlinear computing unit; the nonlinear computing unit operates in an optical amplification mode; the tunable laser and the nonlinear computing unit are connected via an optical fiber.

[0006] Tunable lasers are used to emit optical signals with adjustable wavelength and power.

[0007] The nonlinear arithmetic unit is used to perform nonlinear transformations on optical signals and output nonlinear optical signals; the nonlinear optical signals have a nonlinear activation function relationship with each other.

[0008] In some embodiments, the optical nonlinear computing device further includes an isolator; the isolator is disposed between the tunable laser and the nonlinear computing device and is connected via an optical fiber;

[0009] An isolator is used to eliminate back-reflected light from an optical signal, allowing the optical signal to be transmitted unidirectionally to a nonlinear arithmetic unit.

[0010] In some embodiments, the optical nonlinear computing device further includes: an optical fiber attenuator; the optical fiber attenuator is disposed between the isolator and the nonlinear computing device and is connected via an optical fiber;

[0011] Fiber optic attenuators are used to attenuate optical signals so that the power of the attenuated optical signal is within a preset power range.

[0012] In some embodiments, the operating current of the nonlinear arithmetic unit ranges from 6.5mA to 7.0mA.

[0013] In some embodiments, the power of the optical signal ranges from 2μW to 95μW.

[0014] In a second aspect, this application provides a test system for an optical nonlinear computing device, the test system comprising: an optical nonlinear computing device as described in any embodiment of the first aspect above, and a beam splitter, a circulator, a power meter, a detector, and an analyzer; the beam splitter and the circulator are sequentially disposed between a tunable laser and a nonlinear computing device; the power meter is connected to the beam splitter, and the detector is connected to the circulator; the analyzer is connected to both the power meter and the detector.

[0015] A power meter is used to detect the power value of the optical signal emitted by a tunable laser.

[0016] A detector used to detect the power value of the output optical signal of a nonlinear arithmetic unit;

[0017] The analyzer is used to perform nonlinear curve fitting based on the power values ​​of the optical signal and the output optical signal, generate a fitted curve, and display the fitted curve on the analyzer's display screen.

[0018] Thirdly, this application provides a testing method for an optical nonlinear computing device, which is applied to an analyzer in a testing system for the optical nonlinear computing device as described in the second aspect. The method includes:

[0019] The power value of the optical signal detected by the power meter and the power value of the output optical signal detected by the detector are obtained.

[0020] A nonlinear curve is fitted based on the power values ​​of the optical signal and the output optical signal to generate a fitted curve;

[0021] The fitted curve is displayed on the analyzer's screen.

[0022] In some embodiments, the method further includes:

[0023] Determine whether the fitted curve matches the standard curve;

[0024] When the fitted curve matches the standard curve, the power parameter of the tunable laser is used as the configuration parameter of the tunable laser, and the operating current parameter of the nonlinear arithmetic unit is used as the configuration parameter of the nonlinear arithmetic unit.

[0025] When the fitted curve does not match the standard curve, the tunable laser and / or nonlinear arithmetic unit are readjusted and the test is repeated.

[0026] In some embodiments, retesting involves readjusting the tunable laser and / or the nonlinear arithmetic unit, including:

[0027] Readjust the power parameters of the tunable laser, and / or readjust the operating current parameters of the nonlinear arithmetic unit;

[0028] Based on the power value of the optical signal output by the tunable laser after readjusting the power parameters, and the power value of the optical signal output by the nonlinear arithmetic unit after readjusting the operating current parameters, a nonlinear curve is fitted to generate a new fitted curve.

[0029] In some embodiments, retesting involves readjusting the tunable laser and / or the nonlinear arithmetic unit, including:

[0030] Replace the tunable laser and / or the nonlinear arithmetic unit and retest.

[0031] Fourthly, this application also provides a testing apparatus for optical nonlinear computing devices, the apparatus comprising:

[0032] The acquisition module is used to acquire the power value of the optical signal detected by the power meter and the power value of the output optical signal detected by the detector.

[0033] The fitting module is used to perform nonlinear curve fitting based on the power values ​​of the optical signal and the output optical signal, and generate a fitted curve.

[0034] The display module is used to display the fitted curve on the analyzer's screen.

[0035] Fifthly, this application also provides an analyzer including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0036] The power value of the optical signal detected by the power meter and the power value of the output optical signal detected by the detector are obtained.

[0037] A nonlinear curve is fitted based on the power values ​​of the optical signal and the output optical signal to generate a fitted curve;

[0038] The fitted curve is displayed on the analyzer's screen.

[0039] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0040] The power value of the optical signal detected by the power meter and the power value of the output optical signal detected by the detector are obtained.

[0041] A nonlinear curve is fitted based on the power values ​​of the optical signal and the output optical signal to generate a fitted curve;

[0042] The fitted curve is displayed on the analyzer's screen.

[0043] Fifthly, this application also provides a computer program product, which includes a computer program that, when executed by a processor, performs the following steps:

[0044] The power value of the optical signal detected by the power meter and the power value of the output optical signal detected by the detector are obtained.

[0045] A nonlinear curve is fitted based on the power values ​​of the optical signal and the output optical signal to generate a fitted curve;

[0046] The fitted curve is displayed on the analyzer's screen.

[0047] The aforementioned optical nonlinear computing device, its testing system, and testing method are described. The optical nonlinear computing device includes a tunable laser and a nonlinear computing unit, wherein the nonlinear computing unit operates in optical amplification mode. The tunable laser and the nonlinear computing unit are connected via optical fiber. The tunable laser emits an optical signal with adjustable wavelength and power, and the nonlinear computing unit performs a nonlinear transformation on the optical signal, outputting a nonlinear optical signal. The nonlinear optical signal exhibits a nonlinear activation function relationship with the optical signal. In this optical nonlinear computing device, the nonlinear computing unit operates in optical amplification mode, requiring only a low-power optical signal (microwatt level) from the tunable laser to drive it, yet it can produce a significant nonlinear transformation effect, achieving high-sensitivity optical computing with low power consumption. By adjusting the wavelength of the optical signal emitted by the tunable laser, the input-output response characteristics of the nonlinear computing unit can be flexibly changed, thereby dynamically reconstructing the shape and threshold of the nonlinear activation function relationship, enhancing the device's adaptability and application range. The nonlinear optical signal output by the nonlinear arithmetic unit (NAU) directly exhibits a nonlinear activation function relationship with the input optical signal. This allows the device to function as a nonlinear unit in an optical neural network, directly simulating the required computational functions. Based on the physical mechanism of the NAU in optical amplification mode, its nonlinear transformation process of optical signals inherently possesses high-speed response characteristics, laying the foundation for constructing high-speed optical computing systems. The optical nonlinear arithmetic device consists only of a tunable laser and the NAU connected by optical fiber. Its core architecture is simple, avoiding complex optical path design and debugging, thus facilitating the future realization of large-scale, integrated optical neural networks. Attached Figure Description

[0048] Figure 1 This is one of the structural schematic diagrams of the optical nonlinear computing device in some embodiments;

[0049] Figure 2 This is a second schematic diagram of the structure of the optical nonlinear computing device in some embodiments;

[0050] Figure 3 This is the third schematic diagram of the structure of the optical nonlinear computing device in some embodiments;

[0051] Figure 4 This is a schematic diagram of the test system structure for the optical nonlinear computing device in some embodiments;

[0052] Figure 5 Schematic diagrams of nonlinear transfer curves under different wavelength detuning injections in some embodiments;

[0053] Figure 6 This is one of the flowcharts illustrating the testing method for optical nonlinear computing devices in some embodiments;

[0054] Figure 7 This is a second schematic flowchart of a testing method for an optical nonlinear computing device in some embodiments;

[0055] Figure 8 This is a diagram of the internal structure of the analyzer in some embodiments.

[0056] Explanation of reference numerals in the attached figures:

[0057] Tunable laser 10; nonlinear arithmetic unit 20; isolator 30; fiber optic attenuator 40; beam splitter 50; circulator 60; power meter 70; detector 80; analyzer 90. Detailed Implementation

[0058] In the embodiments of this application, the term "and / or" describes the relationship between associated objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. The character " / " generally indicates that the preceding and following associated objects have an "or" relationship.

[0059] In the embodiments of this application, the term "multiple" refers to two or more, and other quantifiers are similar.

[0060] In the embodiments of this application, the term "at least one" means one or more. For example, at least one of A, B and C can represent six situations: A exists alone, B exists alone, C exists alone, A and B exist simultaneously, A and C exist simultaneously, B and C exist simultaneously, and A, B and C exist simultaneously.

[0061] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0062] Due to the advantages of optical transmission, such as high throughput, low latency, and low energy consumption, optical neural networks (ONNs) are expected to address the challenges of energy consumption and computational efficiency in the current development of artificial intelligence technology, becoming a research hotspot in academia and industry in recent years. In optical neural network computation, nonlinear layers can effectively increase the number of layers in the neural network and help it achieve complex functions. Therefore, how to achieve nonlinear activation in the optical domain is a pressing technical problem in optical neural network computation. In 2019, the team led by Shengwang Tu at the Hong Kong University of Science and Technology proposed the absorption light characteristics of a dark-line two-dimensional magneto-optical trap to achieve nonlinear activation in the optical domain. In 2023, the team led by Peter L. McMahon at Cornell University proposed a spatial optical nonlinear scheme based on an optical saturation amplifier. The saturation nonlinear amplifier acts as a nonlinear transport layer in the ONN neural network, improving the accuracy of the ONN in simple image recognition tasks. Overall, current spatial optical nonlinear schemes are mainly applied to simple tasks such as image preprocessing, with relatively low neural network depth and functional complexity. This is because current nonlinear schemes in spatial optical computation still have the following shortcomings: the nonlinear line type has a low matching degree with commonly used functions such as ReLU and Sigmoid. Nonlinear linearity lacks reconfigurability. The energy threshold for nonlinearity is generally high, and needs to be reduced to the μW level to match the requirements of low-power computing.

[0063] In view of this, embodiments of this application propose an optical nonlinear computing device, a testing system for the optical nonlinear computing device, and a testing method. By using a vertical cavity surface-emitting laser (VCSEL) for optical nonlinear computing in a system design, a nonlinear amplification effect can be achieved, and a nonlinear activation function matching the nonlinear activation function in the neural network can be output.

[0064] It should be noted that the beneficial effects or technical problems solved by the embodiments of this application are not limited to this one, but may also be other implicit or related problems. For details, please refer to the description of the embodiments below.

[0065] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will be described below with reference to the accompanying drawings.

[0066] In some embodiments, such as Figure 1 As shown, an optical nonlinear computing device is provided, which includes a tunable laser 10 and a nonlinear computing unit 20. The tunable laser 10 and the nonlinear computing unit 20 are connected via an optical fiber.

[0067] The nonlinear arithmetic unit 20 operates in optical amplification mode. The tunable laser 10 emits optical signals with adjustable wavelength and power; the nonlinear arithmetic unit 20 performs nonlinear transformations on the optical signals to output nonlinear optical signals. These nonlinear optical signals exhibit a nonlinear activation function relationship with the optical signals. This nonlinear activation function can be a ReLU, Sigmoid, or similar function.

[0068] The aforementioned tunable laser 10 is a light source whose wavelength and power can be precisely controlled. Its wavelength tuning capability is key to achieving "reconfigurable" nonlinearity. The wavelength of the tunable laser 10 is 1550 nm. The power of the optical signal emitted by the tunable laser 10 ranges from 2 μW to 95 μW.

[0069] The aforementioned nonlinear operator 20 is based on a vertical-cavity surface-emitting laser (VCSEL). It utilizes a commercially available or custom-designed VCSEL chip and operates in optical amplification mode by controlling its bias current. In this mode, it functions as a vertical-cavity surface-emitting optical amplifier (VCSOA). Optical amplification mode refers to operating at a subthreshold bias current, close to the current value when the VCSEL is emitting light normally, i.e., the state where the VCSEL is about to emit light or emits weak light. The operating current of the nonlinear operator 20 ranges from 6.5 mA to 7.0 mA.

[0070] The working principle of the optical nonlinear computing device in this application embodiment is as follows: a system design and driving method for using a vertical-cavity surface-emitting laser (VCSEL) for optical nonlinear computing. This device is characterized by operating in a subthreshold state, specifically under subthreshold current injection, and is used as an optical amplification device (VCSOA). Incident light is injected into the VCSOA from the front with a certain wavelength detuned. Different injection powers affect the proportion of carriers participating in stimulated recombination and nonradiative recombination. This process causes changes in the local temperature of the device, resulting in different degrees of shift in the gain peak of the VCSOA. Consequently, the gain obtained by the injected light in the VCSOA changes with the incident power, achieving a nonlinear amplification effect. Specifically, it includes the following three stages:

[0071] The first stage is the system initialization and parameter setting process, including setting the operating mode of the nonlinear arithmetic unit 20 and setting the wavelength detuning of the input light. Specifically: Setting the operating mode of the nonlinear arithmetic unit 20: By applying a bias current lower than its laser emission threshold to the nonlinear arithmetic unit 20, it is made to operate in optical amplification mode (i.e., used as a VCSOA). At this time, the nonlinear arithmetic unit 20 itself does not generate laser light, but it has the ability to amplify optical signals in a specific wavelength band. Its internal gain medium is in a "quasi-equilibrium" state. Setting the wavelength detuning of the input light: Adjusting the tunable laser 10 so that there is a preset difference between the wavelength of its output light and the inherent gain peak wavelength of the nonlinear arithmetic unit 20 in the static state, i.e., "wavelength detuning". This detuning amount is the "key variable" controlling the shape of the final nonlinear function.

[0072] The second stage involves the physical generation process of nonlinear effects, including carrier dynamics and energy conversion, temperature-triggered gain peak shift, and the formation of nonlinear amplification. Specifically: Carrier dynamics and energy conversion: When the optical signal with a specific wavelength detuning emitted by the tunable laser 10 is injected into the nonlinear arithmetic unit 20, it interacts with the carriers in the gain medium. At low power input, most carriers consume energy through nonradiative recombination (such as Auger recombination), which is mainly converted into heat, leading to a local temperature increase in the active region of the device. As the input optical power increases, the proportion of stimulated radiative recombination begins to rise, which consumes carriers and generates coherent photons. Temperature-triggered gain peak shift: The increase in the temperature of the active region causes a change in the bandgap of the semiconductor material, resulting in a redshift in the optical gain spectrum of the nonlinear arithmetic unit 20 (i.e., the gain peak shifts towards longer wavelengths). Nonlinear amplification occurs because the wavelength of the input light is fixed (set by the tunable laser 10), while the gain peak of the nonlinear amplifier 20 shifts with increasing input power. This causes the effective detuning of the input light wavelength relative to the shifting gain peak to continuously change. At a certain critical input power point, the shifting gain peak is closest to the input light wavelength, at which point the net gain provided by the amplifier reaches its maximum. Before this point, the gain gradually increases; after this point, the gain gradually decreases. This dynamic process of first increasing and then decreasing causes the curve of the output nonlinear optical signal's optical power changing with the input optical power to exhibit nonlinear characteristics, such as a typical "S" shape, i.e., the characteristics of the sigmoid function.

[0073] The third stage involves the implementation of the reconfigurable nonlinear function. By changing the two initial conditions set in the first stage, the specific shape of the nonlinear function can be flexibly adjusted. Specifically, different initial wavelength detunings (such as -0.15nm, -0.11nm) directly change the "inflection point" position (nonlinear threshold power) and "saturation value" of the Sigmoid curve. This is equivalent to implementing the switching between various nonlinear activation functions, such as a smooth Sigmoid function and a steeper step-like function, at the hardware level.

[0074] The optical nonlinear computing device provided in this application embodiment has a nonlinear operator 20 operating in optical amplification mode. It only requires a low-power optical signal (microwatt level) from a tunable laser 10 to drive it, yet it can produce significant nonlinear transformation effects, achieving high-sensitivity optical computation with low power consumption. By adjusting the wavelength of the optical signal emitted by the tunable laser 10, the input-to-output response characteristics of the nonlinear operator 20 can be flexibly changed, thereby dynamically reconstructing the shape and threshold of the nonlinear activation function relationship, enhancing the device's adaptability and application range. The nonlinear optical signal output by the nonlinear operator 20 directly exhibits a nonlinear activation function relationship with the input optical signal, enabling the device to function as a nonlinear unit in an optical neural network, directly simulating the required computational function. Based on the physical mechanism of the nonlinear operator 20 in optical amplification mode, its nonlinear transformation process of the optical signal inherently possesses high-speed response characteristics, laying the foundation for constructing a high-speed optical computing system. The optical nonlinear computing device consists only of a tunable laser 10 and a nonlinear operator 20 connected by optical fiber, resulting in a simple core architecture that avoids complex optical path design and debugging, facilitating the future realization of large-scale, integrated optical neural networks.

[0075] In some embodiments, such as Figure 2 As shown, the aforementioned optical nonlinear computing device further includes an isolator 30. The isolator 30 is disposed between the tunable laser 10 and the nonlinear computing unit 20 and is connected via an optical fiber.

[0076] The isolator 30 is used to eliminate the back reflection of the optical signal, so that the optical signal can be transmitted unidirectionally to the nonlinear arithmetic unit 20.

[0077] The optical nonlinear computing device in this embodiment operates as follows: The optical signal generated by the tunable laser 10 first passes through the isolator 30. This component eliminates back-reflected light, allowing the optical signal to be transmitted unidirectionally to the nonlinear computing unit 20. This means it only allows light to flow from the laser to the nonlinear computing unit 20, while blocking any light reflected back from the nonlinear computing unit 20 or other parts. The purified optical signal is injected into the nonlinear computing unit 20 operating in optical amplification mode. After nonlinear transformation, the output nonlinear optical signal is used for subsequent optical processing or detection.

[0078] The optical nonlinear computing device described in this application effectively eliminates back-reflected light from the optical signal by using an isolator 30 positioned between the tunable laser 10 and the nonlinear computing unit 20. This prevents reflected light from returning to the tunable laser 10, thus avoiding interference with the operational stability of the tunable laser 10 and ensuring the reliable operation of the entire optical nonlinear computing device. The isolator 30 ensures unidirectional transmission of the optical signal to the nonlinear computing unit 20, making the optical signal injected into the nonlinear computing unit 20 a controlled forward-propagating light. This eliminates interference from back-reflected light on the operating state of the nonlinear computing unit 20, thereby guaranteeing the accuracy and repeatability of the nonlinear transformation process from optical signal to nonlinear optical signal. By introducing the isolator 30 to form a unidirectional optical path, common unstable factors in optical systems caused by reflection, such as noise and frequency drift, are effectively suppressed. This simplifies system debugging and maintenance, making the optical nonlinear computing device more applicable and robust in practical optical computing environments.

[0079] In some embodiments, such as Figure 3 As shown, the aforementioned optical nonlinear computing device also includes an optical fiber attenuator 40. The optical fiber attenuator 40 is disposed between the isolator 30 and the nonlinear computing unit 20, and is connected via an optical fiber.

[0080] The fiber optic attenuator 40 is used to attenuate the optical signal, ensuring that the power of the attenuated optical signal is within a preset power range. The preset power range is 2μW~95μW.

[0081] The working principle of the optical nonlinear computing device in this embodiment is as follows: the optical signal purified by the isolator 30 enters the optical fiber attenuator 40, the optical fiber attenuator 40 adjusts the intensity of the optical signal so that the power of the attenuated optical signal is within a preset power range, and then the optical signal with the power precisely controlled within the preset power range is injected into the nonlinear computing unit 20 working in optical amplification mode. After nonlinear transformation, the output nonlinear optical signal is used for subsequent optical processing or detection.

[0082] Correspondingly, the sigmoid response curve of the nonlinear operator 20 is exhibited within a preset power range. The fiber optic attenuator 40 ensures that regardless of the laser's output power, the optical power injected into the nonlinear operator 20 is precisely attenuated and fixed within this observable nonlinear "golden range." Furthermore, besides wavelength mistuning, variations in input power also affect the details of the nonlinear curve. By precisely adjusting the injected power through the fiber optic attenuator 40, it can work in conjunction with wavelength tuning to further fine-tune the shape and threshold of the nonlinear activation function.

[0083] The optical nonlinear computing device described in this application embodiment, through an optical fiber attenuator 40 disposed between the isolator 30 and the nonlinear computing unit 20, can precisely attenuate the power of the optical signal transmitted to the nonlinear computing unit 20 to a preset power range, ensuring that the nonlinear computing unit 20 operates in its optimal nonlinear response range, thereby guaranteeing the accuracy and consistency of the nonlinear transformation process from optical signal to nonlinear optical signal. Adjusting the optical signal power input to the nonlinear computing unit 20 through the optical fiber attenuator 40, combined with the wavelength adjustment of the tunable laser 10, provides control methods in both wavelength and power dimensions, enhancing the ability to control the relationship of the nonlinear activation function and the flexibility of reconstruction. The optical fiber attenuator 40 can attenuate the optical signal power to a preset range on the order of μW, enabling the nonlinear computing unit 20 to generate the required nonlinear transformation in an optimal operating state with low power consumption and high sensitivity, which is beneficial for achieving low-power operation of the entire device system.

[0084] In some embodiments, a testing system for optical nonlinear computing devices is also provided, such as... Figure 4 As shown, the test system includes: an optical nonlinear computing device as described in any of the above embodiments, and a beam splitter 50, a circulator 60, a power meter 70, a detector 80, and an analyzer 90. The beam splitter 50 and the circulator 60 are sequentially disposed between the tunable laser 10 and the nonlinear computing device 20. The power meter 70 is connected to the beam splitter 50, the detector 80 is connected to the circulator 60, and the analyzer 90 is connected to both the power meter 70 and the detector 80.

[0085] Among them, the power meter 70 is used to detect the power value of the optical signal emitted by the tunable laser 10; the detector 80 is used to detect the power value of the output optical signal of the nonlinear arithmetic unit 20; and the analyzer 90 is used to perform nonlinear curve fitting based on the power value of the optical signal and the power value of the output optical signal, generate a fitting curve, and display the fitting curve on the display screen of the analyzer 90.

[0086] The working principle of the test system in this application embodiment is as follows: the tunable laser 10 generates an initial optical signal, which passes through the isolator 30 and the fiber optic attenuator 40 in sequence to form a pre-processed optical signal with stable power and purity.

[0087] The optical signal then enters the beam splitter 50 and is split into two paths: the main optical signal continues forward and enters the first port of the circulator 60; the reference optical signal is directed to the power meter 70. The power meter 70 accurately measures and records the power value of this path, which represents the input optical power before it is injected into the nonlinear arithmetic unit 20. This step is crucial, as it provides accurate abscissa data for subsequent nonlinear analysis.

[0088] The main optical signal is routed from the first port of circulator 60 to the second port, and then injected into nonlinear arithmetic unit 20 (i.e., VCSOA operating in subthreshold state). As described in the previous embodiment, nonlinear arithmetic unit 20 performs a nonlinear transformation on the input light in optical amplification mode based on the physical mechanism of carrier-temperature-gain drift. The output optical signal (i.e., the nonlinear optical signal) after the nonlinear transformation is reflected back from nonlinear arithmetic unit 20 and re-enters the second port of circulator 60. Based on its unidirectional transmission characteristics, circulator 60 routes this reflected light to the third port, completing the signal output.

[0089] The nonlinear optical signal output from the third port of the circulator 60 is received by the detector 80. The detector 80 measures its power value, which is the output optical power after nonlinear transformation, and is used as the vertical axis data for analysis.

[0090] The analyzer 90 (such as a computer with dedicated software) synchronously reads each set of corresponding input / output power values ​​from the power meter 70 and detector 80. As the fiber optic attenuator 40 gradually changes the input optical power, the analyzer 90 collects a series of data points (input power, output power). Subsequently, the analyzer 90 uses a preset mathematical model to perform nonlinear curve fitting on these data points to generate a fitted curve. This curve accurately describes the nonlinear transfer function of the device. This mathematical model can be expressed by the following relationship (1):

[0091] (1)

[0092] Where x is the input optical power (i.e., the optical signal power injected into the VCSOA); y is the output optical power (i.e., the optical signal power reflected after nonlinear amplification by the VCSOA); and A, B, C, and D are the parameters to be fitted.

[0093] Finally, the analyzer 90 displays the fitted curve on its screen, visually demonstrating the Sigmoid-type nonlinear characteristics of the optical nonlinear computing device under specific settings (such as current wavelength detuning), as detailed in [reference needed]. Figure 5 The diagram shows the nonlinear transfer curves under different wavelengths of detuned injection.

[0094] The test system for the optical nonlinear computing device described in this application, through the combination of beam splitter 50 and power meter 70, can monitor and record the power value of the optical signal input to nonlinear computing unit 20 in real time; through the combination of circulator 60 and detector 80, it can accurately measure the power value of the output optical signal after nonlinear transformation, thereby providing the analyzer 90 with accurate input-output data pairs and realizing precise quantification of the nonlinear transformation characteristics of the device. The analyzer 90 performs nonlinear curve fitting based on the power values ​​measured by power meter 70 and detector 80, generates the fitting curve, and displays it on the screen, allowing the operator to intuitively observe and analyze the nonlinear activation function relationship and its specific form of the optical nonlinear computing device. By integrating beam splitter 50, circulator 60, power meter 70, detector 80, and analyzer 90, a complete test system is constructed, which can systematically measure, record, and analyze the nonlinear response of optical nonlinear computing devices under different operating parameters, providing a complete test scheme for device performance verification and optimization.

[0095] In some embodiments, a testing method for optical nonlinear computing devices is also provided, the testing method being applied to, for example... Figure 4 The analyzer in the test system of the optical nonlinear computing device shown, such as Figure 6 As shown, the method includes:

[0096] S101, acquire the power value of the optical signal detected by the power meter and the power value of the output optical signal detected by the detector.

[0097] S102 performs nonlinear curve fitting based on the power values ​​of the optical signal and the output optical signal to generate a fitted curve.

[0098] S103 displays the fitted curve on the analyzer's screen.

[0099] In this embodiment, the analyzer can establish a communication connection with the power meter and detector via a data bus (such as GPIB, USB, or Ethernet). During testing, the analyzer can issue commands to the fiber optic attenuator to change the attenuation amount according to a preset sequence (e.g., from 2μW to 95μW in fixed steps). At each attenuation setting point, after the system stabilizes, the analyzer synchronously or sequentially reads the current input optical power value (xi) from the power meter and the corresponding output optical power value (yi) from the detector. The analyzer stores the series of data points (xi, yi) read in memory, forming a complete dataset representing the input-to-output response of the device.

[0100] After acquiring a series of data points (xi,yi), the analyzer can obtain a pre-stored or user-inputted nonlinear fitting mathematical model (i.e., the above relation (1)), call its built-in optimization algorithm (e.g., nonlinear least squares method, such as the Levenberg-Marquardt algorithm), and fit the collected dataset (xi,yi) with the above model. The fitting algorithm iteratively adjusts the values ​​of parameters A, B, C, and D to minimize the sum of squared residuals between the fitted curve and the measured data points, thereby determining a set of optimal parameter values. Using the obtained optimal parameters, the analyzer can calculate a continuous and smooth fitting curve covering the entire input power range on a high-resolution virtual coordinate axis.

[0101] After obtaining the fitted curve, the analyzer calls the graphics library to create a plotting area on the graphical user interface. The fitted curve is plotted as a continuous line (such as a solid line) on the graph. Simultaneously, the collected raw data points (xi, yi) are usually overlaid on the graph as scatter points (such as circles or crosses) for a visual comparison of the fitted effect with the measured data. Finally, axis labels (such as "Input Power (μW)" and "Output Power (μW)", a legend, and a title (such as "Nonlinear Response at -0.11nm Wavelength Detuning") are added to the graph. Figure 5 As shown, different line types are mainly reflected in the power value (nonlinear threshold power) at the S-shaped inflection point. In practical applications, we can select different nonlinear line types according to different needs. It can also display key fitting parameter values ​​(A, B, C, D) and goodness of fit (such as R²), forming the final visualization results. The fitting results of nonlinear functions with different wavelength detunings are shown in Table 1 below:

[0102] Table 1

[0103]

[0104] The method described in this application establishes an accurate input-output power correspondence by synchronously acquiring the power values ​​of the input and output optical signals. This provides a reliable data foundation for the quantitative analysis of nonlinear characteristics, making the characterization of the nonlinear transformation characteristics of optical nonlinear computing devices more precise. By fitting nonlinear curves based on the acquired power values, an accurate mathematical model can be established to describe the nonlinear activation function relationship. Furthermore, the changes in the fitting parameters reflect the variation of the nonlinear response under different operating conditions, providing a theoretical basis for the research and optimization of the device's reconfigurable characteristics. By displaying the fitted curve on the analyzer's screen, the nonlinear response characteristics of the device are presented in an intuitive graphical manner, allowing the operator to directly observe the morphological characteristics of the nonlinear activation function, greatly facilitating the evaluation of device performance and the monitoring of its operating status.

[0105] In some embodiments, such as Figure 7 As shown, the above test method also includes:

[0106] S201, determine whether the fitted curve is consistent with the standard curve.

[0107] The standard curve is a reference model pre-stored in the analyzer. It can be a theoretically calculated ideal sigmoid curve, or it can be the best-fit curve obtained statistically after extensive testing on known qualified "gold sample" devices.

[0108] In this embodiment, the analyzer can compare the parameters (A, B, C, D) of the currently measured fitted curve with the parameters (A0, B0, C0, D0) of the standard curve one by one, and calculate their relative error or absolute error. If the errors of all parameters are within a preset tolerance, they are determined to be consistent. Optionally, within the same input power range, the root mean square error of each point between the measured fitted curve and the standard curve can be calculated. If the root mean square error is lower than a preset threshold, they are determined to be consistent. Optionally, key features of the two curves can be compared, such as nonlinear threshold power (inflection point), saturated output power, and slope in the linear region. If the differences of these feature points are within an allowable range, they are determined to be consistent.

[0109] S202, when the fitted curve matches the standard curve, the power parameter of the tunable laser is used as the configuration parameter of the tunable laser, and the operating current parameter of the nonlinear operator is used as the configuration parameter of the nonlinear operator.

[0110] The configuration parameter can be a single value, multiple values, or a range of values.

[0111] In this embodiment, when the fitted curve matches the standard curve, the analyzer determines that the current operating state of the device under test is excellent and meets the design standards. Then, the power parameters, such as the drive current (or direct output power value) of the tunable laser, are used as configuration parameters for the tunable laser, and the operating current parameters (i.e., its subthreshold bias current) of the nonlinear operator are used as configuration parameters for the nonlinear operator, forming a set of configuration parameters that is saved. The above steps can be iterated until all feasible configuration parameters for the tunable laser and the nonlinear operator are determined. These configuration parameters can be written into the non-volatile memory associated with the optical nonlinear operator, or they can be shipped with the device as its "identity" or "optimal operating point." When the device is integrated into an optical neural network, the system can directly call these configuration parameters to quickly bring it to its optimal operating state.

[0112] S203, when the fitted curve is inconsistent with the standard curve, readjust the tunable laser and / or nonlinear arithmetic unit and retest.

[0113] In this embodiment, the power parameters of the tunable laser and / or the operating current parameters of the nonlinear arithmetic unit can be readjusted. Then, based on the power value of the optical signal output by the tunable laser after the power parameters are readjusted, and the power value of the optical signal output by the nonlinear arithmetic unit after the operating current parameters are readjusted, a nonlinear curve is fitted to generate a new fitted curve, which is then compared with a standard curve to determine the configuration parameters. This process can be iterated until the curves are consistent or the maximum number of iterations is reached.

[0114] Specifically, when the fitted curve differs from the standard curve, the analyzer can automatically generate adjustment instructions based on the direction of the difference between the fitted curve and the standard curve. For example, if the nonlinear threshold is too high, it may instruct the tunable laser to fine-tune its output power parameters (through the drive current), or instruct it to fine-tune the operating current parameters of the nonlinear arithmetic unit to change its initial gain state. After adjustment, the system automatically repeats the above test process, that is, re-acquires data and generates a new fitted curve.

[0115] Optionally, the tunable laser and / or nonlinear operator can be replaced for retesting. Specifically, if the fitted curve consistently fails to match the standard curve after multiple rounds of parameter adjustments, the analyzer may alert the operator that "hardware mismatch or malfunction exists." In this case, the operator should replace the tunable laser and / or nonlinear operator (e.g., replace with a different batch of VCSOA) and then retest the entire test system. This approach is used to diagnose and eliminate inherent defects or batch-to-batch variations in core components.

[0116] In summary, based on all the above embodiments, a testing method for optical nonlinear computing devices is also provided, the method comprising:

[0117] S301, acquire the power value of the optical signal detected by the power meter and the power value of the output optical signal detected by the detector.

[0118] S302 performs nonlinear curve fitting based on the power values ​​of the optical signal and the output optical signal to generate a fitted curve.

[0119] S303 displays the fitted curve on the analyzer's screen.

[0120] S304, determine whether the fitted curve is consistent with the standard curve.

[0121] S305, when the fitted curve matches the standard curve, the power parameters of the tunable laser are used as the configuration parameters of the tunable laser, and the operating current parameters of the nonlinear arithmetic unit are used as the configuration parameters of the nonlinear arithmetic unit.

[0122] S306, When the fitted curve is inconsistent with the standard curve, readjust the power parameters of the tunable laser and / or readjust the operating current parameters of the nonlinear arithmetic unit.

[0123] S307 performs nonlinear curve fitting based on the power value of the optical signal output by the tunable laser after readjusting the power parameters and the power value of the optical signal output by the nonlinear arithmetic unit after readjusting the operating current parameters, generating a new fitted curve.

[0124] S308: When the fitted curve is inconsistent with the standard curve, replace the tunable laser and / or the nonlinear arithmetic unit and retest.

[0125] The methods described in each of the above steps have been described in the foregoing embodiments. For details, please refer to the foregoing descriptions. They will not be repeated here.

[0126] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0127] Based on the same inventive concept, this application also provides a testing apparatus for optical nonlinear computing devices, used to implement the testing method for the aforementioned optical nonlinear computing devices. The solution provided by this apparatus is similar to the implementation described in the above method. Therefore, the specific limitations of one or more testing apparatus embodiments for optical nonlinear computing devices provided below can be found in the limitations of the testing method for optical nonlinear computing devices described above, and will not be repeated here.

[0128] In some embodiments, a testing apparatus for an optical nonlinear computing device is provided, comprising:

[0129] The acquisition module is used to acquire the power value of the optical signal detected by the power meter and the power value of the output optical signal detected by the detector.

[0130] The fitting module is used to perform nonlinear curve fitting based on the power values ​​of the optical signal and the output optical signal, and generate a fitted curve.

[0131] The display module is used to display the fitted curve on the analyzer's screen.

[0132] In some embodiments, the testing apparatus for the above-mentioned optical nonlinear computing device further includes:

[0133] The first determination module is used to determine whether the fitted curve is consistent with the standard curve.

[0134] The second determining module is used to use the power parameters of the tunable laser as the configuration parameters of the tunable laser and the operating current parameters of the nonlinear operator as the configuration parameters of the nonlinear operator when the fitted curve matches the standard curve.

[0135] The adjustment module is used to readjust the tunable laser and / or nonlinear arithmetic unit for retesting when the fitted curve is inconsistent with the standard curve.

[0136] In some embodiments, the adjustment module described above includes:

[0137] The adjustment unit is used to readjust the power parameters of the tunable laser and / or readjust the operating current parameters of the nonlinear arithmetic unit.

[0138] The fitting unit is used to perform nonlinear curve fitting based on the power value of the optical signal output by the tunable laser after readjusting the power parameters and the power value of the optical signal output by the nonlinear arithmetic unit after readjusting the operating current parameters, and to generate a new fitting curve.

[0139] In some embodiments, the adjustment module further includes:

[0140] Replacement unit, used to replace the tunable laser and / or nonlinear arithmetic unit for retesting.

[0141] Each module in the testing device for the aforementioned optical nonlinear computing device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the operations corresponding to each module.

[0142] In some embodiments, an analyzer is provided, which can be a terminal or a server, and its internal structure diagram can be as follows: Figure 8As shown, the analyzer includes a processor, memory, input / output interface, communication interface, display unit, and input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interface. The processor provides computational and control capabilities. The analyzer's memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The analyzer's input / output interface is used for exchanging information between the processor and external devices. The analyzer's communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it implements a testing method for an optical nonlinear computing device. The analyzer's display unit is used to form a visually visible image and can be a display screen, projection device, or virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of the analyzer can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the analyzer housing, or external keyboards, touchpads, or mice, etc.

[0143] Those skilled in the art will understand that Figure 8 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the analyzer to which the present application is applied. A specific analyzer may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0144] In some embodiments, an analyzer is provided, including a memory and a processor, the memory storing a computer program, the processor executing the computer program to perform the following steps:

[0145] The power value of the optical signal detected by the power meter and the power value of the output optical signal detected by the detector are obtained.

[0146] A nonlinear curve is fitted based on the power values ​​of the optical signal and the output optical signal to generate a fitted curve;

[0147] The fitted curve is displayed on the analyzer's screen.

[0148] In some embodiments, when the processor executes a computer program, it further performs the following steps:

[0149] Determine whether the fitted curve matches the standard curve;

[0150] When the fitted curve matches the standard curve, the power parameter of the tunable laser is used as the configuration parameter of the tunable laser, and the operating current parameter of the nonlinear arithmetic unit is used as the configuration parameter of the nonlinear arithmetic unit.

[0151] When the fitted curve does not match the standard curve, the tunable laser and / or nonlinear arithmetic unit are readjusted and the test is repeated.

[0152] In some embodiments, when the processor executes a computer program, it further performs the following steps:

[0153] Readjust the power parameters of the tunable laser, and / or readjust the operating current parameters of the nonlinear arithmetic unit;

[0154] Based on the power value of the optical signal output by the tunable laser after readjusting the power parameters, and the power value of the optical signal output by the nonlinear arithmetic unit after readjusting the operating current parameters, a nonlinear curve is fitted to generate a new fitted curve.

[0155] In some embodiments, when the processor executes a computer program, it further performs the following steps:

[0156] Replace the tunable laser and / or the nonlinear arithmetic unit and retest.

[0157] The computer device provided in the above embodiments has a similar implementation principle and technical effect to the above method embodiments, and will not be described again here.

[0158] In some embodiments, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, performs the following steps:

[0159] The power value of the optical signal detected by the power meter and the power value of the output optical signal detected by the detector are obtained.

[0160] A nonlinear curve is fitted based on the power values ​​of the optical signal and the output optical signal to generate a fitted curve;

[0161] The fitted curve is displayed on the analyzer's screen.

[0162] In some embodiments, when a computer program is executed by a processor, it further performs the following steps:

[0163] Determine whether the fitted curve matches the standard curve;

[0164] When the fitted curve matches the standard curve, the power parameter of the tunable laser is used as the configuration parameter of the tunable laser, and the operating current parameter of the nonlinear arithmetic unit is used as the configuration parameter of the nonlinear arithmetic unit.

[0165] When the fitted curve does not match the standard curve, the tunable laser and / or nonlinear arithmetic unit are readjusted and the test is repeated.

[0166] In some embodiments, when a computer program is executed by a processor, it further performs the following steps:

[0167] Readjust the power parameters of the tunable laser, and / or readjust the operating current parameters of the nonlinear arithmetic unit;

[0168] Based on the power value of the optical signal output by the tunable laser after readjusting the power parameters, and the power value of the optical signal output by the nonlinear arithmetic unit after readjusting the operating current parameters, a nonlinear curve is fitted to generate a new fitted curve.

[0169] In some embodiments, when a computer program is executed by a processor, it further performs the following steps:

[0170] Replace the tunable laser and / or the nonlinear arithmetic unit and retest.

[0171] The computer-readable storage medium provided in the above embodiments has a similar implementation principle and technical effect to the above method embodiments, and will not be described again here.

[0172] In some embodiments, a computer program product is provided, including a computer program that, when executed by a processor, performs the following steps:

[0173] The power value of the optical signal detected by the power meter and the power value of the output optical signal detected by the detector are obtained.

[0174] A nonlinear curve is fitted based on the power values ​​of the optical signal and the output optical signal to generate a fitted curve;

[0175] The fitted curve is displayed on the analyzer's screen.

[0176] In some embodiments, when a computer program is executed by a processor, it further performs the following steps:

[0177] Determine whether the fitted curve matches the standard curve;

[0178] When the fitted curve matches the standard curve, the power parameter of the tunable laser is used as the configuration parameter of the tunable laser, and the operating current parameter of the nonlinear arithmetic unit is used as the configuration parameter of the nonlinear arithmetic unit.

[0179] When the fitted curve does not match the standard curve, the tunable laser and / or nonlinear arithmetic unit are readjusted and the test is repeated.

[0180] In some embodiments, when a computer program is executed by a processor, it further performs the following steps:

[0181] Readjust the power parameters of the tunable laser, and / or readjust the operating current parameters of the nonlinear arithmetic unit;

[0182] Based on the power value of the optical signal output by the tunable laser after readjusting the power parameters, and the power value of the optical signal output by the nonlinear arithmetic unit after readjusting the operating current parameters, a nonlinear curve is fitted to generate a new fitted curve.

[0183] In some embodiments, when a computer program is executed by a processor, it further performs the following steps:

[0184] Replace the tunable laser and / or the nonlinear arithmetic unit and retest.

[0185] The computer program product provided in the above embodiments has a similar implementation principle and technical effect to the above method embodiments, and will not be described again here.

[0186] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0187] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0188] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. An optical nonlinear computing device, characterized in that, The optical nonlinear computing device includes a tunable laser and a nonlinear computing unit; the nonlinear computing unit operates in optical amplification mode; the tunable laser and the nonlinear computing unit are connected via optical fiber. The tunable laser is used to emit optical signals with adjustable wavelength and power. The nonlinear arithmetic unit is used to perform nonlinear transformation on the optical signal and output a nonlinear optical signal; the nonlinear optical signal and the optical signal have a nonlinear activation function relationship.

2. The optical nonlinear computing device according to claim 1, characterized in that, The optical nonlinear computing device further includes: an isolator; the isolator is disposed between the tunable laser and the nonlinear computing device, and is connected via the optical fiber; The isolator is used to eliminate the back reflection of the optical signal, so that the optical signal can be transmitted unidirectionally to the nonlinear arithmetic unit.

3. The optical nonlinear computing device according to claim 2, characterized in that, The optical nonlinear computing device further includes: an optical fiber attenuator; the optical fiber attenuator is disposed between the isolator and the nonlinear computing device, and is connected via the optical fiber; The optical fiber attenuator is used to attenuate the optical signal so that the power of the attenuated optical signal is within a preset power range.

4. The optical nonlinear computing device according to claim 1, characterized in that, The operating current of the nonlinear arithmetic unit ranges from 6.5mA to 7.0mA.

5. The optical nonlinear computing device according to claim 1, characterized in that, The power of the optical signal ranges from 2μW to 95μW.

6. A testing system for an optical nonlinear computing device, characterized in that, The test system includes: an optical nonlinear computing device as described in any one of claims 1-5, and a beam splitter, a circulator, a power meter, a detector, and an analyzer; the beam splitter and the circulator are sequentially disposed between the tunable laser and the nonlinear computing device; the power meter is connected to the beam splitter, and the detector is connected to the circulator; the analyzer is connected to the power meter and the detector respectively. The power meter is used to detect the power value of the optical signal emitted by the tunable laser; The detector is used to detect the power value of the output optical signal of the nonlinear arithmetic unit; The analyzer is used to perform nonlinear curve fitting based on the power value of the optical signal and the power value of the output optical signal, generate a fitting curve, and display the fitting curve on the analyzer's display screen.

7. A testing method for an optical nonlinear computing device, characterized in that, The test method is applied to the analyzer in the test system of the optical nonlinear computing device as described in claim 6, and the method includes: The power value of the optical signal detected by the power meter and the power value of the output optical signal detected by the detector are obtained. A nonlinear curve is fitted based on the power values ​​of the optical signal and the output optical signal to generate a fitted curve; The fitted curve is displayed on the analyzer's screen.

8. The test method according to claim 7, characterized in that, The method further includes: Determine whether the fitted curve is consistent with the standard curve; When the fitted curve matches the standard curve, the power parameter of the tunable laser is used as the configuration parameter of the tunable laser, and the operating current parameter of the nonlinear arithmetic unit is used as the configuration parameter of the nonlinear arithmetic unit. When the fitted curve is inconsistent with the standard curve, the tunable laser and / or the nonlinear arithmetic unit are readjusted and retested.

9. The test method according to claim 8, characterized in that, The retesting of the tunable laser and / or the nonlinear arithmetic unit includes: Readjust the power parameters of the tunable laser, and / or readjust the operating current parameters of the nonlinear arithmetic unit; Based on the power value of the optical signal output by the tunable laser after readjusting the power parameters, and the power value of the optical signal output by the nonlinear arithmetic unit after readjusting the operating current parameters, a nonlinear curve is fitted to generate a new fitted curve.

10. The test method according to claim 8, characterized in that, The retesting of the tunable laser and / or the nonlinear arithmetic unit includes: Replace the tunable laser and / or the nonlinear arithmetic unit and retest.

Images