A Reconfigurable Integrated Optical Computing Core Model

Abstract

The present invention discloses a reconfigurable integrated optical computing core model, comprising: M×N matrix calculation units; each matrix calculation unit includes two dual-waveguide microring resonators, a straight waveguide, and a phase change material heater, and the straight waveguide is made by covering a phase change material on a silicon waveguide; the phase change material heater is used to heat the phase change material to generate a phase change, thereby changing the light absorption of the straight waveguide; the integrated optical computing core model realizes the multiplication operation of an M×N matrix and an N×1 vector through integrated optical methods, where M and N are natural numbers. The present invention uses a phase change material in the computing core, changes the phase state of the phase change material by heating, changes the light absorption of the straight waveguide, and further changes the matrix elements in the computing core, realizing the reconfigurable characteristics of the computing core model. The present invention can be widely applied to the field of integrated optics.

Description

Technical Field

[0001] The present invention relates to the field of integrated optics, and in particular, to a reconfigurable integrated optical computing core model. Background Art

[0002] The development of artificial intelligence has brought a great deal of convenience to our lives. Its foundation is the artificial neural network (ANN), which is a product of simulating biological neural networks. By using various mathematical models to perform a large number of fast and accurate operations for ANN training based on a dataset, complex problem analysis can be carried out after training. However, with the continuous development of AI, there are increasingly high requirements for the data calculation speed and accuracy of ANN, and this growth rate has far exceeded Moore's law of integrated circuits. The mismatch in the development speeds of these two technologies is specifically manifested in that the computing devices for complex neural network training are now large in size and consume an astonishing amount of power. In addition to continuously refining existing integrated circuit technologies, many researchers have also started to pay attention to another completely different field - integrated optics.

[0003] The vector-matrix multiplication operation is one of the basic mathematical operation operations of neural networks, so it has also received extensive research. How to achieve fast and accurate vector-matrix operations based on integrated optics is currently a research hotspot. The optical vector-matrix operation multiplier was initially proposed by J.W. Goodman of Stanford University in the United States in 1978. Once proposed, it has received extensive attention from many scholars. Although it was already recognized at that time that the vector-matrix multiplier based on integrated optics would be very promising in the field of optical computing, due to the limitations of the then technological level, the work in this area had to stagnate.

[0004] Nowadays, with the continuous development of semiconductor manufacturing processes, materials science, on-chip light sources, and photodetectors, some vector-matrix multipliers based on integrated optics have gradually emerged. However, most of them are based on on-chip optical devices such as Mach-Zehnder interferometers or directional couplers, and the modulation methods mostly rely on thermo-optical effects or electro-optical effects. Due to their own volatility, it is very difficult for these vector-matrix multipliers to be scaled up on a large scale, otherwise it will result in extremely high power consumption. At the same time, this type of vector-matrix multiplier also has the disadvantages of large area and slow response speed. Summary of the Invention

[0005] In order to solve at least one of the technical problems existing in the prior art to a certain extent, the purpose of the present invention is to provide a reconfigurable integrated optical computing core model.

[0006] The technical solution adopted by the present invention is as follows:

[0007] A reconfigurable integrated optical computing core model, including M×N matrix calculation units;

[0008] The matrix calculation unit includes two double-waveguide microring resonators, a straight waveguide, and a phase change material heater, and the straight waveguide is made by depositing a phase change material on a silicon waveguide;

[0009] The phase change material heater is used to heat the phase change material to generate a phase change, thereby changing the light absorption of the straight waveguide;

[0010] The integrated optical computing core model realizes the multiplication operation of an M×N matrix and an N×1 vector through integrated optical methods (such as wavelength division multiplexing, demultiplexing, etc.), where M and N are natural numbers.

[0011] Further, the phase change material is Ge2Sb2Te5 (abbreviated as GST), GSST, Sb2S3, Sb2Se3, or In2Se3, etc.

[0012] Further, the waveguides of the two double-waveguide microring resonators are perpendicular to each other but do not cross, and there is no resonance between the two microrings in the same matrix calculation unit.

[0013] Further, the two double-waveguide microring resonators are arranged vertically, and the straight waveguide connects the vertical waveguides of the upper and lower double-waveguide microring resonators:

[0014] The horizontal waveguide of the lowermost double-waveguide microring resonator is located below the microring and serves as the input waveguide of the light wave. The vertical waveguide of the lowermost double-waveguide microring resonator is located on the right side of the microring and serves as the output waveguide of the microring resonator. This output waveguide connects one end of the straight waveguide, and the other end of the straight waveguide connects the vertical waveguide of the uppermost double-waveguide microring resonator. This vertical waveguide is located on the right side of the microring and serves as the input waveguide of the light wave of the uppermost double-waveguide microring resonator. The horizontal waveguide of the uppermost double-waveguide microring resonator is located above the microring and serves as the output waveguide of the light wave of the uppermost double-waveguide microring resonator; the phase change material heater covers the phase change material on the straight waveguide and part of the silicon waveguide and extends to both sides, connecting to the Pd electrodes of the phase change material heater.

[0015] Further, all the matrix calculation units in each row share a common input waveguide and an output waveguide. The input waveguide is located at the bottom of each row, and the output waveguide is located at the top of each row.

[0016] Further, the radii of the double-waveguide microring resonators in the same column of matrix calculation units are the same, and the radii of the double-waveguide microring resonators in the same row of matrix calculation units gradually increase in an arithmetic progression;

[0017] For each double - waveguide microring resonator with a different radius, there is a unique resonant wavelength. The input optical wave that satisfies this wavelength will be coupled from the input waveguide into the microring and then output from the output waveguide; the input optical wave that does not satisfy the resonant wavelength will propagate backward along the input waveguide to the next matrix calculation unit.

[0018] Furthermore, the material of the phase - change material heater is indium tin oxide (ITO), and the electrode material of the phase - change material heater is palladium (Pd).

[0019] The phase - change material heater covers all of the phase - change material and part of the silicon waveguide.

[0020] Furthermore, after the phase - change material is heated by the phase - change material heater, if the temperature exceeds the glass transition temperature and is less than the melting point, the phase - change material will change from a completely amorphous state to a completely crystalline state; if the temperature exceeds the melting point, the phase - change material will change from a completely crystalline state to a completely amorphous state.

[0021] Different elements of the summary matrix of the computing core are characterized by using different phase states of the phase - change material; the phase - change material is heated by the phase - change material heater to cause a phase change of the phase - change material, thereby achieving reconfigurability, that is, the elements in the computing core matrix can be arbitrarily changed.

[0022] Among them, the elements of the matrix are any positive numbers greater than or equal to 0 and less than or equal to 1.

[0023] Furthermore, the input vector of the integrated optical computing core model is the laser with different wavelengths generated by a continuous - wave laser. The wavelengths respectively correspond to the resonant wavelengths of the double - waveguide microring resonators in each row of the matrix calculation unit, and at the same time, the laser power of each input respectively corresponds to the elements of the matrix input vector.

[0024] Furthermore, the phase - change material includes multiple intermediate states other than the completely amorphous state and the completely crystalline state, and when the phase - change material is in different phase states, there are different absorption loss coefficients w j .

[0025] Furthermore, when the wavelength of the phase - change material is 1555 nm, the refractive index in the completely crystalline state is 6.392 + 1.976i, and the refractive index in the completely amorphous state is 4.153 + 0.099i.

[0026] The beneficial effects of the present invention are as follows: The present invention uses a phase - change material in the computing core. By heating, the phase state of the phase - change material is changed, the absorption of light by the straight waveguide is changed, and then the matrix elements in the computing core are changed, realizing the reconfigurable characteristic of the computing core model. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following introduces the accompanying drawings of the related technical solutions in the embodiments of the present invention or the prior art. It should be understood that the accompanying drawings in the following introduction are only for clearly expressing some embodiments of the technical solutions in the present invention for convenience. For those skilled in the art, without creative efforts, other accompanying drawings can also be obtained based on these drawings.

[0028] Figure 1 is the complete structural schematic diagram of the computing unit (excluding the phase change material heater) in the embodiment of the present invention;

[0029] Figure 2 is the transmittance-crystallization degree curve of the phase change material GST used in the computing unit in the embodiment of the present invention;

[0030] Figure 3 is the structural schematic diagram of the first row in the 5×5 matrix (excluding the phase change material heater) in the embodiment of the present invention;

[0031] Figure 4 is the complete structural schematic diagram of the 5×5 matrix (excluding the phase change material heater) in the embodiment of the present invention;

[0032] Figure 5 is the output current curve of the 5×5 matrix in the embodiment of the present invention;

[0033] Figure 6 is the complete structural schematic diagram of the computing unit including the phase change material heater in the embodiment of the present invention;

[0034] Figure 7 is the temperature-time curve when GST is crystallized in the embodiment of the present invention;

[0035] Figure 8 is the temperature-time curve when GST is amorphous in the embodiment of the present invention.

[0036] Figure 1 Reference numeral: 1. Thin layer of phase change material GST covering the waveguide. Detailed implementation manners

[0037] The following details the embodiments of the present invention. The examples of the embodiments are shown in the accompanying drawings, where the same or similar reference numerals represent the same or similar elements or elements with the same or similar functions throughout. The embodiments described below by referring to the accompanying drawings are exemplary and are only used to explain the present invention and should not be construed as a limitation to the present invention. For the step numbers in the following embodiments, they are only set for the convenience of elaboration and illustration, and no limitation is imposed on the order between the steps. The execution order of each step in the embodiments can be adaptively adjusted according to the understanding of those skilled in the art.

[0038] In the description of the present invention, it should be understood that regarding the orientation description, such as the orientation or positional relationship indicated by up, down, front, back, left, right, etc. is based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation to the present invention.

[0039] In the description of the present invention, the meaning of several is one or more, the meaning of multiple is more than two, greater than, less than, exceeding, etc. are understood as not including the recited number, and above, below, within, etc. are understood as including the recited number. If there is a description of first and second, it is only for the purpose of distinguishing technical features and cannot be understood as indicating or implying relative importance or implicitly indicating the quantity of the indicated technical features or implicitly indicating the sequence relationship of the indicated technical features.

[0040] In the description of the present invention, unless otherwise clearly defined, terms such as set, install, connect, etc. should be understood in a broad sense, and those skilled in the art can reasonably determine the specific meanings of the above terms in the present invention in combination with the specific content of the technical solution.

[0041] Aiming at the problems existing in the prior art, the present invention relates to a reconfigurable integrated optical computing core model based on phase change materials and dual waveguide microring resonators. The reconfigurable optical computing core is a matrix model containing M×N computing units (i.e., matrix computing units). Each computing unit is composed of two dual waveguide microring resonators with the same radius, a phase change material-silicon hybrid straight waveguide, and a phase change material heater for reconstructing the computing unit. The computing units in each column have dual waveguide microring resonators with the same radius, and the radius of the dual waveguide microring resonators in each row of computing units gradually increases. For a dual waveguide microring resonator with a specific radius r, an input optical wave with a specific wavelength λ r will be coupled into the microring and output from the other waveguide. Based on this mechanism, each row of the matrix can simultaneously input N different wavelength signals determined by the radius of its dual waveguide microring resonators. Further assume that the powers of the N different wavelength input optical waves are P j , where j = 1, 2, 3... N. At the same time, the input optical wave is output to the phase change material-silicon hybrid straight waveguide section. At this time, different absorption loss coefficients w j will be generated according to the different phase states of the phase change material, where j = 1, 2, 3... N. Here, a voltage can be applied to the phase change material heater to increase its temperature and heat the phase change material, thereby changing the phase state of the phase change material and further changing the absorption loss coefficient w j , that is, realizing the reconstruction of the computing core. Finally, the output optical wave power will become P j ×wj , and then through the demultiplexing operation, the output optical waves of different wavelengths are separated and then passed through the photodetector array corresponding to their wavelengths, and the final current value output by the entire computing core can be obtained, that is, the final result of the vector-matrix operation.

[0042] The above integrated optical computing core model will be explained in detail below with reference to the accompanying drawings.

[0043] Figure 1 is the most basic computing unit in the matrix (excluding the phase change material heater). As shown in the figure, it is composed of two dual-waveguide microring resonators and a section of GST-silicon hybrid straight waveguide. The black square represents that GST is in the fully crystalline state. If the radius of the microring is r, when an optical wave with a wavelength of λ r is input from port 1 (assuming that λ r is the resonance wavelength of the dual-waveguide microring resonator in this computing unit), at this time, due to resonance, the input optical wave will be coupled from the bottom straight waveguide into the lower microring. Then the optical wave propagates counterclockwise in the microring. When it approaches the straight waveguide on the right side in the microring, it will be coupled with it, causing the optical wave in the microring to be coupled into the straight waveguide on the right side. At this time, it is equivalent to the optical wave being output from port 2. Then the optical wave continues to propagate along the straight waveguide and enters the GST-silicon hybrid straight waveguide section. When passing through the GST thin layer, if GST is in the crystalline state at this time, as described above, the imaginary part of its refractive index is very large, and the imaginary part of the refractive index of a material is also called the extinction factor, which affects the absorption of the optical wave by the material. The greater the extinction factor, the stronger the absorption of the optical wave by the material, and vice versa. Therefore, according to the different phase states of GST, the absorption degree of the GST-silicon hybrid straight waveguide to the input optical wave will be different, that is, its transmittance will change, and then it is reflected as different degrees of decrease in optical power. Then the optical wave continues to propagate along the waveguide to port 3, where the same resonance as described above occurs, and finally the optical wave is output from port 4. This process is characterized by the mathematical expression: P out = P in ·w. Where P in is the optical wave power input at port 1, P out is the optical wave power output at port 4, and w is the transmittance of the GST-silicon hybrid waveguide.

[0044] Therefore, using such a computing unit can complete a scalar calculation, as shown by the mathematical expression P out = P in ·w. And through precise phase change control, the crystallization degree of the phase change material GST can theoretically make the transmittance of the GST-silicon hybrid waveguide, that is, the multiplier w in the operation formula, satisfy: 0 ≤ w ≤ 1. The transmittance of the GST-silicon hybrid waveguide under different crystallization degrees is as Figure 2 shown.

[0045] Figure 3 It is a schematic diagram of the first row structure of a 5×5 matrix as an embodiment. In the figure, the radii of the dual-waveguide microring resonators in the 5 computing units gradually increase in an arithmetic progression. Assuming that the optical wave wavelengths input at this time respectively correspond to the resonant wavelengths of the dual-waveguide microring resonators in the five computing units of this row, which are λ1, λ2, λ3, λ4, λ5 in sequence, and the corresponding optical wave powers are P1, P2, P3, P4, P5, then the total optical wave energy input is P in = P1 + P2 + P3 + P4 + P5. When the input optical wave propagates along the bottom straight waveguide to the first microring, only the input optical wave with the resonant wavelength λ1 corresponding to its radius r1 is coupled into the microring, and the input optical waves with other wavelengths continue to propagate backward. The resonant conditions of the remaining microrings are similar to that of the first microring. After the input optical wave is coupled by the following microrings, it is respectively output to the GST-silicon hybrid straight waveguide sections of their corresponding computing units for the aforementioned scalar operations, and then coupled to the top straight waveguide through the upper microrings and output from the leftmost end. Due to the use of the wavelength division multiplexing characteristic of light, the optical wave power finally output is equal to the sum of the optical powers output by each computing unit, that is, there is the following mathematical expression: P out = P1·w1 + P2·w2 + P3·w3 + P4·w4 + P5·w5, where w1, w2, w3, w4, w5 respectively correspond to the transmittances of the GST-silicon hybrid straight waveguides in the five computing units.

[0046] It is not difficult to see that Figure 3 the structure shown can complete a vector-vector multiplication operation. Taking the power P of the input optical wave j as a column vector, and taking the transmittance w of the GST-silicon hybrid straight waveguide in each computing unit j as a row vector, then the final total output power is the operation result of the vector-vector multiplication, and the mathematical expression is as follows:

[0047]

[0048] Figure 4 It is a schematic diagram of the complete structure of a 5×5 matrix as an embodiment. When understanding Figure 3After understanding the working principle of the single-row matrix structure shown, it is not difficult to see that the principle of performing vector-matrix multiplication in this embodiment is very intuitive, that is, to perform the aforementioned vector-vector multiplication operation 5 times simultaneously. At the same moment, light waves with 5 different wavelengths are input to all rows of the matrix. These 5 different wavelengths respectively correspond to the resonant wavelengths of the dual-waveguide microring resonators in the 5 computing units of each row. However, at this time, the transmittance of the GST-silicon hybrid straight waveguide in the computing unit of each row is different, and it is specifically determined according to the matrix to be calculated at that time. Finally, the output light waves of all rows of the matrix pass through the photodetector array, and the output power is converted into an output current value to obtain the result of the vector-matrix multiplication operation. The mathematical expression of the calculation process is as follows:

[0049]

[0050] The elements in the above matrix respectively correspond to the transmittance of the GST-silicon hybrid waveguide in the computing unit. In theory, the transmittance, that is, the matrix elements, can satisfy: 0 ≤ w ij ≤ 1. It should also be noted that the photodetector array used here is assumed to be a PIN photodiode detector array, and its responsivity is 1.

[0051] From Figure 4 it can be seen that the number of phase change materials in the crystalline state (black squares) in the computing unit of each row is different, decreasing sequentially from top to bottom. The matrix of the embodiment Figure 4 is characterized by the mathematical expression:

[0052]

[0053] At this time, it is assumed that the input optical wave power of each row of the matrix is 5W, then it is equivalent to having the mathematical operation:

[0054]

[0055] Combined with Figure 5 the matrix current output diagram shown, it can be found that the result obtained by the operation of the embodiment is basically the same as the result obtained by the mathematical operation.

[0056] Figure 6 is the computing unit in the computing core that includes a phase change material heater. Next, the reconfigurable characteristics of the computing core will be described. As Figure 6 shown, when electrode 1 is grounded and voltages of 5.5V with a duration of 20ns and 3.5V with a duration of 180ns are applied to electrode 2, since Joule heat is generated, the temperature of the heater rises, and then the temperature is transferred to the GST below the heater, finally making its temperature exceed the glass transition temperature and be less than the melting point. Therefore, it is converted from the amorphous state to the crystalline state. The temperature-time curve of crystallization is as Figure 7As shown. Similarly, by grounding electrode 1 and applying a voltage of 7V for 20ns to electrode 2, the temperature of GST can exceed the melting point, thereby converting from the crystalline state to the amorphous state. The temperature-time curve of the amorphization is as shown in Figure 8 As shown. Therefore, by integrating with electronic devices, the phase state of GST can be controlled by electrical signals, that is, the element values of the matrix in the computing core can be controlled, thereby achieving the reconfigurable characteristic, and the speed reaches the microsecond level.

[0057] The on-chip optical computing matrix based on phase change materials and dual-waveguide microring resonators provided by the present invention can be fabricated using a variety of materials, such as Si, SiN, and SOI, etc. There are also various choices for phase change materials, such as GST, GSST, Sb2S3, Sb2Se3, and In2Se3, etc. This vector-matrix multiplier has ultra-low static power consumption, extremely fast computing speed, and at the same time has the reconfigurable characteristic at the microsecond level. Compared with the existing research, the almost zero static power consumption and the reconfigurable characteristic at the microsecond level are very big advantages. The entire structure has the characteristics of small device volume, simple fabrication, compatibility with existing manufacturing processes, fast computing speed, strong scalability, and fast reconfigurability.

[0058] In summary, compared with the prior art, the present embodiment has the following advantages and beneficial effects:

[0059] (1) The computing core described in the present invention is based on CMOS process, which matches well with the existing semiconductor manufacturing technology and does not require the development of new manufacturing technology.

[0060] (2) The computing core described in the present invention uses GST, so it has the reconfigurable characteristic. Changing the phase state of GST in the computing unit means changing the matrix elements in the computing core, enabling the computing core to be applied to calculations in various scenarios.

[0061] (3) GST can stably maintain its phase state for a long time without additional energy consumption, that is, it has non-volatility. In the vector-matrix multiplier, the static power consumption is almost zero, greatly reducing the overall power consumption of the device.

[0062] (4) Due to the use of non-volatile phase change materials, the static power consumption of the computing core is close to zero. Therefore, the computing core described in the present invention has very high scalability.

[0063] (5) The computing core described in the present invention uses a voltage-controlled phase change material heater, enabling the computing core to cooperate well with the electrical system and be integrated with electrical devices. At the same time, the heating speed of the phase change material heater can reach the microsecond level, making the phase change process of GST, that is, the matrix element reconstruction time of the computing core, also reach the microsecond level.

[0064] It should be understood that various parts of the present invention can be implemented by hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, any one or a combination of the following techniques well known in the art can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), and the like.

[0065] In the above description of this specification, the descriptions referring to the terms "one embodiment / example", "another embodiment / example", or "certain embodiments / examples", etc. mean that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0066] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the claims and their equivalents.

[0067] The above has specifically described the preferred embodiments of the present invention, but the present invention is not limited to the above embodiments. Those skilled in the art can also make various equivalent deformations or substitutions without departing from the spirit of the present invention, and these equivalent deformations or substitutions are all included within the scope defined by the claims of this application.

Claims

1. A reconfigurable integrated optical computing core model, characterized in that It includes M×N matrix calculation units; The matrix calculation unit includes two double-waveguide microring resonators, a straight waveguide, and a phase-change material heater. The straight waveguide is made by covering a phase-change material on a silicon waveguide; The phase-change material heater is used to heat the phase-change material to generate a phase change, thereby changing the light absorption of the straight waveguide; The integrated optical computing core model realizes the multiplication operation of an M×N matrix and an N×1 vector through an integrated optical method, where M and N are natural numbers; The two double-waveguide microring resonators are arranged vertically, and the straight waveguide connects the vertical waveguides of the upper and lower double-waveguide microring resonators; The radii of the double-waveguide microring resonators in the matrix calculation units of the same column are the same, and the radii of the double-waveguide microring resonators in the matrix calculation units of the same row gradually increase in the form of an arithmetic progression; For each double-waveguide microring resonator with a different radius, there is a unique resonant wavelength, and the input optical wave that satisfies this wavelength will be coupled into the microring from the input waveguide and then output from the output waveguide; The input optical wave that does not satisfy the resonant wavelength will propagate backward along the input waveguide to the next matrix calculation unit.

2. The reconfigurable integrated optical computing core model according to claim 1, wherein The phase-change material is Ge2Sb2Te5.

3. A reconfigurable integrated optical computing core model according to claim 1, characterized in that The waveguides of the double-waveguide microring resonators are perpendicular to each other but do not cross, and there is no resonance between the two microrings in the same matrix calculation unit.

4. A reconfigurable integrated optical computing core model according to claim 1, characterized in that, All the matrix calculation units in each row share a common input waveguide and an output waveguide. The input waveguide is located at the bottom of each row, and the output waveguide is located at the top of each row.

5. A reconfigurable integrated optical computing core model according to claim 1, characterized in that, The material of the phase-change material heater is indium tin oxide (ITO), and the electrode material of the phase-change material heater is metal palladium (Pd); The phase-change material heater covers all of the phase-change material and part of the silicon waveguide.

6. A reconfigurable integrated optical computing core model according to claim 1, wherein, After the phase-change material is heated by the phase-change material heater, if the temperature exceeds the glass transition temperature and is less than the melting point, the phase-change material will change from a completely amorphous state to a completely crystalline state; if the temperature exceeds the melting point, the phase-change material will change from a completely crystalline state to a completely amorphous state; Different elements of the matrix in the computing core summary are characterized by using different phases of the phase-change material; the phase-change material is heated by the phase-change material heater to cause a phase change of the phase-change material, thereby realizing reconfigurability, that is, the elements of the matrix in the computing core can be arbitrarily changed; Among them, the elements of the matrix are any positive numbers greater than or equal to 0 and less than or equal to 1.

7. A reconfigurable integrated optical computing core model according to claim 1, characterized in that, The phase change material includes a plurality of intermediate states other than the completely amorphous state and the completely crystalline state, and different absorption loss coefficients w exist when the phase change material is in different phase states. j .

8. The reconfigurable integrated optical computing core model according to claim 1, characterized in that The input vector of the integrated optical computing core model is the laser with different wavelengths generated by a continuous-wave laser. The wavelengths respectively correspond to the resonant wavelengths of the double-waveguide microring resonators in each row of matrix calculation units, and at the same time, the power of each input laser respectively corresponds to the elements of the matrix input vector.

Images