Method of testing a photonic computing system

Abstract

The application relates to the field of photonic computing, and discloses a testing method for a photonic computing system, which comprises the following steps: generating a first optical signal according to a first state; performing photoelectric conversion and quantization on the first optical signal to obtain a second electrical signal, wherein the second electrical signal is an electrical signal after quantization; and keeping the first state unchanged and measuring a characteristic parameter of the second electrical signal for multiple times.

Description

Technical Field

[0001] This invention relates to the field of semiconductors, and more specifically, to a testing method for a photonic computing system. Background Technology

[0002] In photonic computing systems, it is sometimes necessary to obtain certain characteristic parameters of light, such as intensity, phase, or polarization. One common method is to convert the optical signal into an electrical signal through photoelectric conversion, and then process the electrical signal to indirectly obtain the characteristic parameters of the light by obtaining the parameters of the electrical signal.

[0003] The inventors have discovered that nonlinear phenomena in the photoelectric conversion process, the precision of devices in the computing system, and the quantization of electrical signals after photoelectric conversion can all lead to the final electrical signal failing to accurately reproduce the true optical signal parameter information, or failing to accurately reproduce the parameter information of the electrical signal corresponding to the true optical signal, such as the parameter information of the electrical signal before quantization. Summary of the Invention

[0004] One embodiment of the present invention proposes a testing method for a photonic computing system, comprising generating a first optical signal according to a first state; performing photoelectric conversion and quantization on the first optical signal to obtain a second electrical signal, wherein the second electrical signal is a quantized electrical signal; and keeping the first state unchanged, measuring the characteristic parameters of the second electrical signal multiple times.

[0005] In some implementations, generating the first optical signal includes generating the first optical signal based on first data.

[0006] This includes the step of amplifying the electrical signal after photoelectric conversion using an amplification circuit.

[0007] In some implementations, the first optical signal is generated by an optical coding unit.

[0008] In some embodiments, the parameters of the second electrical signal are selected from at least one of current parameters and voltage parameters.

[0009] In some implementations, the method includes averaging the characteristic parameters of the second electrical signal.

[0010] In some implementations, the gain of the amplifier circuit is adjusted such that at least two different measurement results exist in the multiple measurements of the characteristic parameters of the second electrical signal.

[0011] In some implementations, parameter information of the first optical signal is obtained by averaging the characteristic parameters of the second electrical signal.

[0012] In some implementations, this includes storing characteristic parameters of the second electrical signal measured multiple times.

[0013] In some implementations, parameter information of the first optical signal is obtained by averaging the characteristic parameters of the second electrical signal.

[0014] In some implementations, this includes storing characteristic parameters of the second electrical signal measured multiple times.

[0015] In some embodiments, the optical coding unit includes a modulator.

[0016] The embodiments of the present invention can more accurately measure the parameter information of the optical signal to be measured in the photonic computing system, or obtain the real information of the related electrical signal, such as the parameter information of the electrical signal before quantization.

[0017] Various aspects, features, advantages, etc., of the embodiments of the present invention will be specifically described below in conjunction with the accompanying drawings. These aspects, features, advantages, etc., will become clearer from the following detailed description in conjunction with the accompanying drawings. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of a test method for a photonic computing system according to an embodiment of the present invention.

[0019] Figure 2 This is a schematic diagram of a photonic computing system according to an embodiment of the present invention.

[0020] Figure 3a , Figure 3b This is a schematic diagram of a photonic computing system according to an embodiment of the present invention.

[0021] Figure 4 This is a schematic diagram of a photonic computing system according to an embodiment of the present invention. Detailed Implementation

[0022] To facilitate understanding of the various aspects, features, and advantages of the technical solution of this invention, the invention will be described in detail below with reference to the accompanying drawings. It should be understood that the various embodiments described below are for illustrative purposes only and are not intended to limit the scope of protection of this invention.

[0023] The terminology used in this disclosure is for the purpose of describing particular embodiments only and is not intended to limit the disclosure. As used in this disclosure, the singular forms “a,” “an,” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms “comprising” and / or “including” as used in this specification designate the presence of the stated features, integrals, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or groups thereof. As used in this disclosure, the term “and / or” includes any and all combinations of one or more of the associated listed items, and the phrase “at least one of A and B” means only A, only B, or both A and B. In this disclosure, a chip may include a bare die. In this disclosure, features of one embodiment may also be applied and appropriately incorporated as features of other embodiments described in this disclosure.

[0024] In one exemplary embodiment of the present invention, a testing method for a photonic computing system is proposed, such as... Figure 1 As shown, the method includes:

[0025] S100 generates a first optical signal according to the first state;

[0026] S200 performs photoelectric conversion and quantization on the first optical signal to obtain a second electrical signal, which is a quantized electrical signal;

[0027] S300 keeps the first state unchanged and measures the characteristic parameters of the second electrical signal multiple times.

[0028] Figure 2 A photonic computing system 100 is shown, which includes a portion for performing computations in the photonic domain, specifically including a light coding unit 110 and a photonic computing unit 130.

[0029] In some implementations, the computing system may include a replication unit.

[0030] Figure 3a A photonic computing system 100 is shown, which includes a portion for performing computation in the photonic domain. Specifically, it includes a photonic encoding unit 110, a copying unit 120, and a photonic computing unit 130.

[0031] Combined Figure 1In step S100, the initial optical signal can be encoded by the optical encoding unit 110. For example, the initial optical signal can be encoded according to the first data to become a first optical signal representing the first data, thereby generating the first optical signal. The optical encoding unit 110 may include a modulator to modulate the initial optical signal with a first electrical signal, wherein the first electrical signal may be the first data from the memory or provided by an external test circuit.

[0032] In step S200, photoelectric conversion can be achieved by a photoelectric conversion unit, such as a photodiode, wherein the photonic computing unit may include the photoelectric conversion unit.

[0033] In step S300, the input state of the optical coding unit can be kept unchanged by keeping the first data unchanged, that is, by keeping the first state unchanged, so that the characteristic parameters of the second electrical signal can be measured multiple times under the same input state of the optical coding unit.

[0034] For example, the optical encoding unit 110 may include multiple modulators, which can encode light according to data. Figure 3b Modulators 101a and 101b in the optical encoding unit are shown. The optical encoding unit is optically connected to the replication unit, which may include one or more replication modules. Exemplary replication modules include beam splitters that can split the light input to the replication module into two or more beams, i.e., perform a "replication" operation to generate an optical copy. This optical copy can serve as the first optical signal. Figure 3b The diagram illustrates replication modules 102a and 102b within a replication unit. The computing system also includes a photonic computing unit configured to receive a first optical signal, which corresponds to first data. The photonic computing unit includes multiple weighting modules, which can also be referred to as multiplication modules, exemplified by... Figure 3b The weighting modules 103a, 103b, 103c, 103d, 103e, and 103f shown herein each correspond to a numerical value and are used to perform a multiplication operation between a numerical value corresponding to one optical signal and a numerical value corresponding to one weighting module. For example, the numerical value represented by the weighting module can be changed by an electrical signal. The output signal of the replication module does not necessarily have the same amplitude as the input signal. For instance, if an optical beam splitter (optical power separator) is used to uniformly separate the input signal power between two output signals, each of the two output signals will have a power equal to or less than 50% of the input signal power.

[0035] In some cases, the replication module or optical encoding unit can be omitted, and the optical signal carrying the data can be directly input into the photonic computing unit through some optical ports (such as optical coupling structures); in addition, the optical signal output by the optical encoding unit can also be directly input into the photonic computing unit.

[0036] In some implementations, photonic integrated circuits can be used to implement optical computing units and optical coding units. Photonic integrated circuits can be manufactured using semiconductor processes to form photonic integrated circuit chips. A photonic integrated circuit chip may include photonic devices such as optical coupling structures, waveguides, photoelectric conversion units, electro-optic conversion units, and light sources. The number of various photonic devices can be configured as needed, and there may be one or more. For example, the electro-optic conversion unit may include a modulator to convert electrical signals into optical signals. For example, the optical coupling structure can be used to optically couple with a laser or optical fiber to input or output optical signals to the photonic integrated circuit chip; for example, optical fiber can be used for optical signal input and output. The optical coupling structure may include grating couplers, end-face couplers, etc. Waveguides can be used to propagate optical signals, serving as channels for information transmission. For example, the photoelectric conversion unit may include a photodetector for converting optical signals into electrical signals; the photodetector may include, for example, a photodiode.

[0037] An optical encoding unit can encode light so that the optical signal represents a numerical value. The encoded optical signal can be input into a photonic computing unit as a data source for computation. For example, the optical encoding unit can encode n numerical values ​​to generate n corresponding optical signals. Exemplarily, the photonic computing unit includes multiple weighting modules, and the n optical signals can be input into the corresponding n weighting modules for computation. The optical encoding unit may include multiple modulators that modulate the light based on the electrical signals representing the numerical values, thereby encoding and generating the initial optical signal. In a photonic computing system, in addition to the aforementioned photonic computing unit, other photonic computing units may be included, such as a second photonic computing unit, a third photonic computing unit, etc. The computation modules in different photonic computing units may differ, and the functions they implement may also differ.

[0038] In addition to the aforementioned weighting modules (used to correspond to multiple values), the photonic computing unit may also include other structures or photonic computing modules capable of performing photonic computing, such as a photonic computing module that uses an MZI (Mach Zehnder interferometer) for calculation.

[0039] In some implementations, generating the first optical signal includes generating the first optical signal based on first data.

[0040] In some embodiments, the method includes a step of amplifying the electrical signal after photoelectric conversion using an amplification circuit. The amplification circuit may include, for example, a transimpedance amplifier (TIA).

[0041] In some embodiments, generating the first optical signal includes generating the first optical signal through an optical coding unit.

[0042] In some embodiments, the parameters of the second electrical signal are selected from at least one of current parameters and voltage parameters.

[0043] In some implementations, the optical signal in the photonic computing system is converted from light to light and then quantized. For the quantized signal, an analog-to-digital converter (ADC) is used during the quantization process. The quantized value is discrete, being an integer multiple of the least significant bit (LSB) of the ADC, such as -5LSB, -2LSB, 0LSB, 1LSB, 2LSB, 3LSB. This results in a certain discrepancy with the true value; that is, the quantized value often does not reflect the true value before quantization.

[0044] In addition, noise is often present in photonic computing systems. When the noise is very small or there is no noise, some data that are similar before quantization will have the same result after quantization, and it will be impossible to distinguish the difference between them when outputting after quantization.

[0045] The inventors have discovered that by introducing appropriate noise or adjusting the gain of the amplifier circuit to amplify the noise already present in the system, effectively perturbing the signal before quantization, different signal outputs can be generated after quantization. Therefore, by obtaining multiple quantized signal outputs, statistical analysis can be performed on these multiple quantized signals to obtain more accurate information about the signal before quantization. For example, averaging the multiple quantized signals can yield more accurate information about the electrical signal before quantization, or its corresponding actual optical signal.

[0046] In some implementations, the characteristic parameters of the second electrical signal are measured multiple times, for example, 50 times, 100 times, or 200 times.

[0047] In some implementations, the method includes averaging the characteristic parameters of the second electrical signal.

[0048] In some implementations, the gain of the amplifier circuit is adjusted such that at least two different measurement results exist among the multiple measurements of the characteristic parameters of the second electrical signal. This allows even small noise in the original computing system to be amplified, resulting in different measurement results after quantization. This makes it easier to obtain the parameter information before quantization based on the average of multiple quantized measurements.

[0049] In some embodiments, the parameter information of the first optical signal is obtained by averaging the characteristic parameters of the second electrical signal. The average value of the obtained second electrical signal is considered to more accurately reproduce the true parameters of the first optical signal, and the parameter information of the first optical signal can be obtained based on the specific processes of photoelectric conversion, amplification, and other processing of the first optical signal, as well as the physical relationships of signal conversion.

[0050] In some implementations, a photonic computing system may include an electronic computing unit, thereby performing a portion of the computation in the electronic domain.

[0051] For example, Figure 4 In comparison Figure 3a In some embodiments, the photonic computing system 100 may further include an electronic computing unit 140. Figure 4 In the middle, a computing system includes: an optical encoding unit 110 configured to generate a first optical signal according to a first data encoding; a photonic computing unit 130 configured to perform a first calculation on the first optical signal in a photonic domain and obtain a second signal representing the calculation result; and an electronic computing unit 140 configured to obtain the second signal and perform a second calculation in an electronic domain based on the second signal.

[0052] In some embodiments, the second signal is an optical signal, and the electronic computing unit includes a photoelectric conversion unit configured to convert the second signal into an electrical signal for performing the second calculation in the electronic domain.

[0053] In some embodiments, the second signal is an electrical signal, and the photonic computing unit includes a photoelectric conversion unit configured to photoelectrically convert the optical result of the first calculation in the photonic domain to obtain a second signal for performing the second calculation in the electronic domain.

[0054] Those skilled in the art should understand that the above-disclosed embodiments are merely implementations of the present invention and should not be construed as limiting the scope of the patent protection claimed in this invention. Equivalent variations made according to the embodiments of the present invention are still within the scope of the claims of the present invention.

Claims

1. A testing method for a photonic computing system, comprising: A first optical signal is generated based on the first state; The first optical signal is photoelectrically converted and quantized to obtain a second electrical signal, which is a quantized electrical signal. While keeping the first state unchanged, measure the characteristic parameters of the second electrical signal multiple times; The test method further includes: After the photoelectric conversion, the electrical signal after the photoelectric conversion is amplified by an amplification circuit; The gain of the amplifier circuit is adjusted so that at least two different measurement results exist in the multiple measurements of the characteristic parameters of the second electrical signal.

2. The method as described in claim 1, wherein, The generation of the first optical signal includes generating the first optical signal based on the first data.

3. The method as described in claim 1, wherein, The generation of the first optical signal includes generating the first optical signal through an optical coding unit.

4. The method of claim 1, wherein, The parameters of the second electrical signal are selected from at least one of current parameters and voltage parameters.

5. The method of claim 1, wherein, This includes the step of averaging the characteristic parameters of the second electrical signal.

6. The method of claim 5, wherein, This includes obtaining parameter information of the first optical signal by averaging the characteristic parameters of the second electrical signal.

7. The method of claim 6, wherein, This includes storing characteristic parameters of the second electrical signal measured multiple times.

8. The method of claim 3, wherein, The optical coding unit includes a modulator.

Images