JP2006023505A - Arithmetic circuit - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an arithmetic circuit with which arithmetic processing, such as summing and summing of products and so on is carried out, without being influenced by a diffraction limit of light by finding a photophysical phenomenon distinct among quantum dots arranged in nanometer region. <P>SOLUTION: An input side quantum dot group 20, consisting of a plurality of quantum dots 12 different in size so as to make signal light beams different in frequency corresponding to input signals be respectively supplied thereto and respective excitons be excited to respective energy levels, corresponding to the frequencies of the supplied signal beams, and an output side quantum dot 13, having resonance energy levels into which excitons are transformed from respective quantum dots 12 constructing the input side quantum dot group 20, corresponding to resonance with the respective energy levels and generating output light beams with wavelengths longer than those of the respective signal light beams as output signals, corresponding to the energy emitted from the resonance energy levels, are formed on a substrate 11. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an arithmetic circuit using quantum dots, particularly applied to the fields of nanoscale optical communication networks and optical measurement.

  With the recent development of semiconductor microfabrication technology, a semiconductor element having a fine structure up to a size at which a quantum mechanical effect appears noticeably has been realized (for example, see Non-Patent Document 1). As semiconductor elements utilizing this quantum mechanical effect, for example, HBT (Hetero-junction Bipolar Transistor) and quantum well lasers have been put into practical use. In addition, nanoscale quantum dots that take advantage of the particle properties of electrons by controlling single electrons using quantum mechanical effects have attracted attention.

  Quantum dots are formed by using the above-described semiconductor microfabrication technology to form a box with a potential that is so fine that it gives three-dimensional quantum confinement to excitons. Utilizing this exciton confinement system, the energy levels of carriers in the quantum dot become discrete, and the density of states sharpens in a delta function. Single-electron memories that use light absorption between the sharpened states of this quantum dot and single-electron transistors that turn on / off single electrons that enter and exit the quantum dot have already been studied. Nanoscale manipulation is being realized.

M. Ohtsu, K. Kobayashi, T. Kawazoe, S. Sangu, T. Yatsui, IEEE J. Sel. Top. Quant. Electron., To be published Vol8.No4 2002 July-Aug, P839-P862 M.J.O'Mahony, D. Simeonidou, D. K. Hunter, A. Tzanakaki, IEEE Commn. Mag. 39, 128 (2001)

  By the way, in order to meet future demands for large-capacity information processing, realization of nanoscale arithmetic circuits, delay circuits, etc. that can perform arithmetic processing, information processing, delay processing, etc. without being governed by the diffraction limit of light Is desired.

  However, there is a problem that quantum fluctuations occur when an attempt is made to realize such a nanoscale circuit with an electronic device. There is a problem that miniaturization is limited. For example, optical packet switching (see, for example, Non-Patent Document 2) that can process main data without converting it to the electrical domain has also been proposed. However, there are various problems to be solved for such switching technology. For example, a large number of specific waveguide devices must be mounted for collation with the packet header, and the required solid-state device is large in scale and difficult to integrate.

  In addition, in recent years, a content addressable memory (CAM) in which an input signal (content) is input to a memory system and an in-memory address that matches the input signal is returned has been proposed, and Routing, Translation Lookup is proposed. -Applied to Side Buffer (TLB), Image Processing, Data Compression (LZW), and the like.

  In order to realize such a CAM mechanism, an evaluation mechanism for the entire data of a plurality of bits is important. However, in an existing electronic circuit such as a MOS transistor, there is a problem that the power consumption becomes large and the realization is difficult. In addition, as an optical means for realizing this, a method of adding a photocurrent with a Position sensitive diode (PSD) and a method of performing a global operation on a processor after converting to an electric region using a light receiving element array are proposed. However, all have difficulty in integration.

  In addition, as an optical data matching mechanism in such a CAM mechanism, there is a system using optical domain code division multiplexing, but it is necessary to prepare a waveguide device for each code individually, which is also difficult to integrate. There is.

  Furthermore, when trying to realize a CAM mechanism with a quantum device, there is a problem that it is difficult to ensure long-term reliability in coherence. For this reason, a practical nanoscale circuit itself has not yet been devised.

  Therefore, the present invention has been devised in view of the above-described problems. A unique photophysical phenomenon is found between quantum dots arranged in the nanometer region, and the sum is not controlled by the light diffraction limit. An object of the present invention is to provide an arithmetic circuit capable of performing arithmetic processing such as arithmetic and product-sum operations.

  In order to solve the above-described problem, the present inventor supplies signal lights having different frequencies corresponding to the input signal, and excitons are excited to energy levels in accordance with the frequency of the supplied signal light. The input side quantum dot group consisting of a plurality of quantum dots of different sizes and excited from each quantum dot constituting the input side quantum dot group through the substrate according to the resonance with each energy level A quantum dot on the output side which has a resonance energy level into which a child is injected and generates output light having a longer wavelength than each signal light as an output signal according to the energy emitted from the resonance energy level. Invented the arithmetic circuit formed.

  In other words, the arithmetic circuit according to the present invention is an arithmetic circuit that generates an output signal represented as the sum of a plurality of input signals, and is supplied with a dielectric substrate and signal lights having different frequencies corresponding to the input signal. An input-side quantum dot group in which a plurality of quantum dots having different sizes are formed on a substrate so that excitons are excited to energy levels according to the frequency of the supplied signal light, and There is a resonance energy level in which excitons are injected from each quantum dot constituting the input side quantum dot group through the substrate in accordance with resonance with the energy level, and the energy released from the resonance energy level Correspondingly, an output-side quantum dot that generates output light having a longer wavelength than each signal light is provided.

  That is, the arithmetic circuit according to the present invention is an arithmetic circuit that generates an output signal expressed as a sum of products of a plurality of input signals, and includes a dielectric substrate, a first signal light corresponding to the input signal, and a second signal. A first quantum dot having a first energy level to which excitons are excited in response to the supplied first signal light, and the same level as the first energy level; A second quantum dot having a second energy level of which excitons are excited to the second energy level in response to the supplied second signal light, and first and second energies A third quantum dot having a third energy level lower than the level, so that the first quantum dot and the second quantum dot are symmetrical with each other via the third quantum dot. Multiple input-side quantum dot glues formed on the substrate And a resonance energy level in which excitons are injected from each input-side quantum dot group in response to resonance with the third energy level in the third quantum dot, and is emitted from the resonance energy level. An output-side quantum dot that generates output light as an output signal according to energy.

  That is, the arithmetic circuit according to the present invention is an arithmetic circuit that generates an output signal expressed as a sum of products of a reference signal with respect to a digitized input signal, and includes a dielectric substrate and a first corresponding to each bit of the input signal. 1 signal light is supplied, and second signal light corresponding to each bit of the reference signal to be integrated with each bit is supplied and excited according to the supplied first signal light. A first quantum dot having a first energy level at which a child is excited and a second energy level at the same level as the first energy level, Accordingly, a second quantum dot in which excitons are excited at the second energy level, and a third quantum having a third energy level lower than the first and second energy levels. Consisting of dots, via a third quantum dot Resonance between a plurality of input-side quantum dot groups formed on the substrate such that the first quantum dot and the second quantum dot are symmetrical with each other, and the third energy level of the third quantum dot Output side quantum dots having resonance energy levels from which excitons are injected from each input side quantum dot group, and generating output light as an output signal according to the energy emitted from the resonance energy levels. With dots.

  That is, the arithmetic circuit according to the present invention includes an arithmetic circuit for obtaining a similarity of each digitized input signal to each reference signal, a plurality of arithmetic processing means to which the input signal and different reference signals are respectively supplied, Similarity identifying means for identifying similarity based on an output signal output from the arithmetic processing means is provided. Each arithmetic processing means includes a dielectric substrate and a first signal light corresponding to each bit of the input signal. The second signal light corresponding to each bit of the reference signal to be integrated with each bit is supplied, and excitons are excited in accordance with the supplied first signal light. The first quantum dot having the first energy level, the second energy level that is the same level as the first energy level, and the second quantum level according to the supplied second signal light. Energy level A second quantum dot in which excitons are excited and a third quantum dot having a third energy level lower than the first and second energy levels. A plurality of input-side quantum dot groups formed on the substrate so that the first quantum dot and the second quantum dot are symmetrical to each other through the third quantum dot; and a third energy level in the third quantum dot; Output side that has a resonance energy level in which excitons are injected from each input side quantum dot group according to the resonance of the output, and generates output light as an output signal according to the energy emitted from the resonance energy level Quantum dots.

  The present invention finds a unique photophysical phenomenon between quantum dots arranged in the nanometer region, and supplies signal lights of different frequencies corresponding to the input signal, respectively, and according to the frequency of the supplied signal light. An input-side quantum dot group consisting of a plurality of quantum dots that are different in size from each other so that excitons are excited to energy levels, and an input-side quantum dot group is configured according to resonance with each energy level. Each quantum dot has a resonance energy level in which excitons are injected through the substrate, and generates output light having a longer wavelength than each signal light as an output signal according to the energy emitted from the resonance energy level. The quantum dots on the output side are formed on the substrate.

  As a result, in the present invention, it becomes possible to provide a nanoscale circuit that can perform arithmetic processing without being controlled by the diffraction limit of light, and can meet the demand for future large-capacity information processing. It becomes.

  Hereinafter, the best mode for carrying out the present invention will be described in detail with reference to the drawings.

First, the arithmetic circuit 1 using quantum dots to which the present invention is applied will be described. The arithmetic circuit 1 is an arithmetic circuit that generates an output signal expressed as the sum of a plurality of input signals. For example, as shown in FIG. 1, the arithmetic circuit 1 is made of a conductive material such as NaCl, KCl, or CaF _2. Substrate 11, one output-side second quantum dot 13 formed on the surface of substrate 11, and a plurality of first quanta discretely formed around second quantum dot 13 And a quantum dot group 20 composed of dots 12. In the following description, the case where the quantum dot group 20 is configured by two first quantum dots 12a and 12b will be described as an example.

  Each of the quantum dots 12a and 12b and the second quantum dot 13 constituting the quantum dot group 20 is based on a discrete energy level formed by confining excitons three-dimensionally. Control). Between the quantum dots 12 and 13, the energy level of the carriers in the quantum dots becomes discrete due to the exciton confinement system, and the state density can be sharpened in a delta function.

  Signal light as near-field light is independently supplied to the quantum dots 12a and 12b constituting the quantum dot group 20. Here, the near-field light supplied to the first quantum dots 12 a is signal light A, and the near-field light supplied to the first quantum dots 12 b is signal light B. The signal light A and the signal light B are supplied to the quantum dots 12a and 12b through the plasmon waveguides 31 and 32 formed on the substrate 11, respectively. However, the present invention is not limited to this case. The quantum dots 12a and 12b may be supplied via a near-field optical probe (not shown) adjacent to the quantum dots 12a and 12b.

  The second quantum dot 13 generates output light according to excitons injected from the first quantum dot 12a or the first quantum dot 12b. A plasmon waveguide 33 for propagating the generated output light and outputting it to the outside is provided around the second quantum dot 13.

  That is, in the arithmetic circuit 1, the signal light A and the signal light B are independently supplied to the first quantum dots 12a and 12b. The second quantum dot 13 generates output light in accordance with the signal lights A and B supplied to the quantum dots 12a and 12b constituting the quantum dot group 20.

  Incidentally, each quantum dot 12 and 13 consists of material systems, such as CuCl, GaN, or ZnO. Incidentally, when the material system constituting each quantum dot 12, 13 is CuCl, these are configured as a cube, and when the material system constituting each quantum dot 12, 13 is GaN or ZnO, these are: Configured as a sphere or disk.

  Each of these quantum dots 12 and 13 can be formed on the substrate 11 by using the following Bridgman method. In the case of using the above-mentioned CuCl as a material system constituting each quantum dot 12, 13, CuCu powder and NaCl powder are first mixed and melted at a temperature of about 800 ° C. Next, the molten mixed powder is suspended in a furnace having a temperature gradient in the vertical direction, and moved up and down in the furnace at a speed of several mm / h, thereby creating a temperature gradient inside the mixed powder gradually. To crystallize. When heat treatment is performed at a temperature of about 200 ° C. for several minutes to several tens of minutes, a NaCl crystal including CuCl quantum dots 12 and 13 can be produced. Incidentally, in this Bridgman method, the size of the generated quantum dots 12 and 13 can be freely controlled by changing the heat treatment temperature and heat treatment time, and these can be formed side by side in a region of 100 nm or less. Become.

  Each of these quantum dots 12 and 13 may be further formed on the substrate 11 based on a molecular epitaxy (MBE) growth method, and the formation position of the quantum dots with high precision using near-field light CVD. You may control.

The quantum confinement level E ( _nx , _ny , _nz ) in each quantum dot 12,13 is expressed by the following formula (1) when the mass of the particle is m and the side length of the quantum dot is L. Defined by
E (n _x , n _y , n _z ) = h ² / 8π ² m (π / L) ² (n _x ² + _ny ² + n _z ² ) (1)
Based on this equation (1), E ( _nx , _ny , _nz ) of each quantum dot 12,13 is calculated. Here, when the side length ratio of the quantum dot 12a, the quantum dot 12b, and the quantum dot 13 is about 1: √2: 2, as shown in FIG. 2, the quantum level in the first quantum dot 12a Is equal to (1,1,1) and E (222) when the quantum level in the second quantum dot 13 is (2,2,2). When the quantum level in the first quantum dot 12b is (1,1,1) and the quantum level in the second quantum dot 13 is (2,1,1). E (211) becomes equal. That is, the quantum level (1, 1, 1) of the first quantum dot 12a and the quantum level (1, 1, 1) of the first quantum dot 12b are the quantum levels in the second quantum dot 13, respectively. (2,2,2), (2,1,1) and the excitation energy levels of excitons resonate with each other. Actually, in order to cause resonance between them, the signal light A having the wavelength λ1 corresponding to the quantum level (1,1,1) in the first quantum dot 12a, the quantum level in the first quantum dot 12b, By supplying the signal light B having the wavelength λ2 corresponding to (1, 1, 1), excitons can be excited to the quantum level.

  If the exciton is excited to the quantum level (1,1,1) in the first quantum dot 12a by supplying the signal light A having the wavelength λ1, the quantum level (1,1,1, Resonance occurs between 1) and the quantum level (2, 2, 2) in the second quantum dot 13. As a result, excitons existing at the quantum level (1, 1, 1) in the first quantum dot 12a move to the quantum level (2, 2, 2) of the second quantum dot 13, and further It moves to the quantum level (1, 1, 1) of the second quantum dot 13. As a result, the excitons apparently move from the first quantum dot 12 a to the second quantum dot 13. The excitons that have moved to the quantum level (1, 1, 1) of the second quantum dot 13 emit light therefrom. Light emitted from the lower quantum level (1, 1, 1) of the second quantum dot 13 is extracted as output light as an output signal, and its wavelength λ3 is the wavelength λ1 of the signal light, Compared with λ2, it becomes longer. This is because the size of the second quantum dot 13 is larger than that of each of the quantum dots 12a and 12b. Accordingly, by selectively detecting only the output light having the wavelength λ3, the light intensity of the output light can be accurately measured without mixing other wavelength components.

  That is, by forming the quantum dots 12 and 13 having different side length ratios on the substrate 11, the quantum levels based on the formula (1) can be made substantially equal, and resonance is caused between them. Thus, excitons can be injected from the quantum dot 12 having a small volume into the quantum dot 13 having a large volume. In other words, excitons can be transmitted between the quantum dots by making the volumes (sizes) different from one another.

  For this reason, by using the transmission principle of the exciton, the signal light A and B corresponding to the first quantum dots 12a and 12b having a small volume are supplied to excite the excitons in the respective quantum levels. This is transmitted to the second quantum dot 13 having a large volume. The second quantum dot 13 generates output light by emitting the transmitted excitons to a lower level, and transmits this to the outside as an output signal.

  Incidentally, the light intensity of the output light from the second quantum dot 13 is governed by the exciton emission amount to the lower level in the second quantum dot 13, and the exciton emission amount is the quantum dot 12 a, Depends on the amount of excitons transmitted from 12b. That is, when the signal lights A and B are respectively supplied to all the first quantum dots 12 a and 12 b, the amount of excitons excited by that amount is increased and transmitted to the second quantum dots 13. As the amount of excitons increases, the light intensity of the output light based on these emissions increases. On the other hand, when the signal lights A and B are not supplied to all the first quantum dots 12a and 12b, the amount of excitons to be excited correspondingly decreases, and the second quantum dots 13 As the amount of excitons transmitted to the light source decreases, the light intensity of the output light based on these emissions decreases. This means that the light intensity of the output light can be changed by controlling the supply state of the signal lights A and B to the quantum dots 12a and 12b.

  The relationship of the light intensity of the output light with respect to the supply state of the output light can also be explained from the simulation result of the time variation of the population shown in FIG. In FIG. 3 (a), excitons are present at both the quantum level (1, 1, 1) of the first quantum dot 12a and the quantum level (2, 2, 2) of the second quantum dot 13. The population in the state where the exciton has moved to the quantum level (1,1,1) in the second quantum dot 13 when it exists (in the biexciton system) is indicated by a solid line. The population in the state in which the exciton does not transit to the quantum level (1,1,1) in the second quantum dot 13 and remains in the quantum level (2,2,2) as the upper level. It is indicated by a broken line.

  The reason why the population indicated by the broken line is oscillating is that if the quantum level (2, 2, 2) as the upper level in the second quantum dot 13 is occupied by excitons, the Pauli In other words, the exciton flow from the first quantum dots 12a is restricted by the exclusion rule. In such a case, excitons travel back and forth between the first quantum dot 12a and the second quantum dot 13 until the quantum level (2, 2, 2) is vacant (so-called chapter movement ( nutation) will be repeated, and this will appear in the form of population vibration. Incidentally, even when this nutation occurs, the exciton finally moves to the quantum level (1, 1, 1) as the lower level in the second quantum dot 13.

  In addition, in FIG. 3B, in addition to the above-described biexciton system, the second quantum dot for each so-called single exciton system in which excitons exist only in the second quantum dot 13 in the initial state. 13 shows the time change of the quantum level (1, 1, 1) as the lower level in FIG. Physically, the output signal is the result of time integration of the output light generated accompanying the exciton transition to the quantum level (1, 1, 1) in the second quantum dot 13. This corresponds to the time integration of the population shown in FIG. When this population is actually integrated on the order of 0 to 5 ns, it can be seen that the ratio of the output signals of the biexciton system and the monoexciton system is about 1.86: 1. This indicates that the light intensity of the output light increases as the number of excitons traveling between the second quantum dot 13 and the first quantum dot 12 increases.

  4 and 5 show the experimental results of the light intensity change of the output light from the second quantum dot 13. As shown in FIG. 4A, when three first quantum dots 12c, 12d, and 12e having a size ratio of about 1: 3: 4 are discretely formed around the second quantum dot 13. When only the signal light C having a wavelength λ (= 325 nm) that can excite the excitons in the first quantum dots 12c is irradiated, the excited excitons move to the second quantum dots 13, where Output light of wavelength λ (= 384 nm). The light intensity of the output light is expressed as the peak intensity at the photon energy W1 corresponding to the wavelength λ (= 384 nm) as shown in FIG.

  Similarly, when only the signal light D having the wavelength λ (= 376 nm) that can excite the excitons in the first quantum dots 12d is irradiated, the light intensity of the output light from the second quantum dots 13 corresponding thereto is As shown in FIG. 4B, it is expressed as the peak intensity at the photon energy W1 corresponding to the wavelength λ (= 384 nm). Similarly, when only the signal light E having a wavelength λ (= 381.3 nm) that can excite excitons in the first quantum dots 12e is irradiated, the light intensity of the output light from the second quantum dots 13 corresponding thereto Is represented as the peak intensity at photon energy W1 corresponding to the wavelength λ (= 384 nm), as shown in FIG.

  As can be seen from the experimental results in FIG. 4B, when any one of the signal lights C, D, and E is irradiated, the light intensity of the output light is almost equal. In other words, when the number of supplied signal lights is one, it means that output light having a certain level of light intensity can be emitted regardless of the wavelength of the signal light.

  On the other hand, when two signal lights E are simultaneously supplied to the first quantum dots 12 in addition to the signal light C, the excitations excited in the first quantum dots 12c and 12e, respectively. The child moves to the second quantum dot 13 from which output light having a wavelength λ (= 384 nm) is emitted. The light intensity of this output light is supplied as shown in FIG. Compared with the case where the number of signal lights is one, it is about twice.

  Similarly, when three signal lights D and E in addition to the signal light C are simultaneously supplied to the first quantum dots 12, they are excited in the first quantum dots 12c, 12d, and 12e, respectively. The excitons move to the second quantum dot 13 and output light having a wavelength λ (= 384 nm) is emitted therefrom, and the light intensity of this output light is supplied as shown in FIG. Compared with the case where the number of signal lights is one, the number of signal lights is about three times.

  In such a case, when the state of light collection from the first quantum dot 12 to the second quantum dot 13 is evaluated in the spatial region by scanning the near-field probe, an intensity profile as shown in FIG. 6 is obtained. . That is, it can be seen that optical signals are collected in an area finer than the wavelength by the near field.

  As can be seen from these experimental results, the light intensity of the output light increases almost linearly according to the number of supplied signal lights. This suggests that the light intensity of the output light is expressed as a sum obtained by adding the number of signal lights to be supplied, and if this can be taken out as an output signal, the arithmetic circuit 1 is connected to the number of signal lights. It is also possible to operate as a so-called summing circuit for summing.

  When the signal light supplied to the arithmetic circuit 1 represents each bit of the digital signal, the state in which each signal light is supplied is set to H level, and the state in which each signal light is not supplied is not present Is set to the L level, the light intensity of the output light increases according to the number of the supplied signal lights. By identifying this, the number of the H level signal lights can be counted.

In such a case, as shown in FIG. 1, the supply of the signal lights A and B to the first quantum dots 12a and 12b is linked with a digital signal to be counted, thereby producing an output signal having a light intensity corresponding to this. Produces the output light. For example, when a 2-bit digital signal (a ₀ , a ₁ ) expressed in binary is obtained, A0 = a ₁ + a ₀ (where a _{1 to} a ₀ take one of values 1 or 0). _Defines the value of a ₀ with signal light A and the value of a ₁ with signal light B.

Here, when the value of a ₀ is 1, the signal light A of wavelength λ1 is supplied (set to H level). On the other hand, when the value of a ₀ is 0, the supply of the signal light A having the wavelength λ1 is stopped (set to L level). In the first quantum dot 12 to which the signal light A is supplied, excitons are excited to the quantum level (1, 1, 1), and the exciton becomes the quantum level of the second quantum dot 13 ( Since the light is emitted by moving to (2, 2, 2) and further transitioning to the lower level, it contributes to the light intensity of the output signal as output light. On the other hand, in the first quantum dot 11 to which the signal light A is not supplied, the exciton is not excited to the quantum level (1, 1, 1). It does not move to the quantum level (2, 2, 2) of the quantum dot 13, and this does not contribute to the light intensity of the output signal as output light.

Similarly, when the value of a ₁ is 1, the signal light B having the wavelength λ2 is supplied. On the other hand, when the value of a ₁ is 0, the supply of the signal light B having the wavelength λ2 is stopped. In the first quantum dot 12 to which the signal light B having the frequency λ2 is supplied, excitons are excited to the quantum level (1, 1, 1). Since the light is emitted by moving to the level (2,1,1) and further transitioning to the lower level, it contributes to the light intensity of the output signal as the output light. On the other hand, in the first quantum dot 12 to which the signal light B is not supplied, the exciton is not excited to the quantum level (1, 1, 1). It does not move to the quantum level (2, 1, 1) of the quantum dot 13, and this does not contribute to the light intensity of the output signal as output light.

For example, when A0 is obtained for a digital signal (00 in binary notation) represented by a ₁ = 0 and a ₀ = 0, signal light is supplied to any first quantum dot 12. The amount of excitons transmitted to the second quantum dot 13 is minimized, and the light intensity L0 of the output light based on these emissions is minimized.

On the other hand, when A0 is obtained for a digital signal (10 in binary notation) represented by a ₁ = 0 and a ₀ = 1, the signal light A is supplied only to the first quantum dot 12a. Thus, since the amount of excitons transmitted to the second quantum dot 13 increases, the light intensity L1 of the output light based on these emissions also increases.

When A0 is obtained for a digital signal (11 in binary notation) represented by a ₁ = 1, a ₀ = 1, the signal light A is supplied to the first quantum dot 12a, and the first The signal light B is supplied to the second quantum dot 12b, and the amount of excitons transmitted to the second quantum dot 13 is the case of a digital signal represented by a ₁ = 0 and a ₀ = 1. Since it increases in comparison, the light intensity L2 of the output light based on these emissions is also nearly twice as large as L1.

  That is, by investigating in advance the relationship between the number of signals supplied to the light intensity of these output lights, the number of H levels in each bit of the digital signal can be summed only by identifying the light intensity of the output light. The calculated result can be obtained.

  As described above, in the arithmetic circuit 1 to which the present invention is applied, signal light having different wavelengths corresponding to each bit of the digital signal to be summed is supplied, and excitons are excited in accordance with the supplied signal light. The plurality of first quantum dots 12 and the second quantum dots 13 having resonance energy levels in which excitons are injected from these quantum levels are formed on the substrate.

  Therefore, by controlling the supply state of the signal light corresponding to the bits of the digital signal to be summed, the total energy flowing from the first quantum dot 12 to the second quantum dot 13 can be controlled, Since the light intensity of the output light as the output signal to be emitted is controlled by this, the number of signals to be supplied is counted, in other words, a summing circuit representing the sum of bit 1 of the digital signal is realized. Can do. If there are N bits of a digital signal that can be input, the arithmetic circuit 1 can also obtain A0 expressed by the following equation (2).

  In particular, in the arithmetic circuit 1 to which the present invention is applied, it is possible to perform summation between nanometer-sized quantum dots without being controlled by the diffraction limit of light. Can also be realized.

  Note that the present invention is not limited to the arithmetic circuit 1 as a summing circuit for obtaining the sum of a plurality of input signals as described above. For example, an output signal expressed as a product sum of a plurality of input signals is generated. You may apply to the arithmetic circuit 2 to do.

  In the arithmetic circuit 2, the same components and elements as those of the arithmetic circuit 1 described above are denoted by the same reference numerals, and the description thereof is omitted here.

For example, as shown in FIG. 7, the arithmetic circuit 2 includes a substrate 11 made of a conductive material such as NaCl, KCl, or CaF ₂ and a second output-side second formed on the surface of the substrate 11. Quantum dots 13 and a quantum dot group 20 composed of a plurality of first quantum dots 12 formed discretely around the second quantum dots 13. In the following description, the case where the quantum dot group 20 is configured by two first quantum dots 12a and 12b will be described as an example.

  Two third quantum dots 16a and a fourth quantum dot 17a are formed in the vicinity of the first quantum dot 12a, and a third quantum dot 16b is formed in the vicinity of the first quantum dot 12b. Two of the fourth quantum dots 17b are formed. Hereinafter, the third quantum dot 16 and the fourth quantum dot 17 are collectively referred to as a calculation unit 10.

  Each quantum dot 16 and 17 which comprises the calculating part 10 controls a single electron (exciton) based on the discrete energy level formed by confining an exciton three-dimensionally. Between the quantum dots 16 and 17, the exciton confinement system makes the energy levels of the carriers in the quantum dots discrete, and the state density can be sharpened in a delta function.

  Signal light as near-field light is independently supplied to each of the quantum dots 16 and 17 constituting the arithmetic unit 10. Here, the near-field light supplied to the third quantum dot 16a is referred to as signal light SA1, and the near-field light supplied to the fourth quantum dot 17a is referred to as signal light SA2. The near-field light supplied to the third quantum dot 16b is signal light SB1, and the near-field light supplied to the fourth quantum dot 17b is signal light SB2. The signal light SA1 and the signal light SA2 are respectively supplied to the quantum dots 16a and 17a via the plasmon waveguides 41 and 42 formed on the substrate 11, and the signal light SB1 and the signal light SB2 are formed on the substrate 11. The quantum dots 16b and 17b are supplied to the quantum dots 16b and 17b via the plasmon waveguides 43 and 44, respectively. However, the present invention is not limited to this case. It may be supplied via a field light probe.

  That is, in the arithmetic circuit 2, the signal light SA1, the signal light SB1, the signal light SA2, and the signal light SB2 are independently supplied to the third quantum dots 16a and 16b and / or the fourth quantum dots 17a and 17b. Supplied. The first quantum dots 12a and 12b use the second excitons according to the signal light SA1, the signal light SB1, the signal light SA2, and the signal light SB2 supplied to the quantum dots 16 and 17 constituting the arithmetic unit 10, respectively. Move to quantum dot 13.

  Here, when attention is paid to the first quantum dot 12a and the third quantum dot 16a and the fourth quantum dot 17a formed in the vicinity thereof, the state in which the signal lights SA1 and SA2 are supplied is set to the H level. If the state in which the signal lights SA1 and SA2 are not supplied is set to the L level, the arithmetic circuit 2 turns the exciton movement on and off in accordance with the signal lights SA1 and SA2, thereby causing an H level signal value, It functions as a circuit for performing a logical operation on the L-level signal value.

  Incidentally, the quantum dots 16 and 17 are made of a material system such as CuCl, GaN, or ZnO, and the manufacturing method is also based on the Bridgman method, the molecular epitaxy (MBE) growth method, or the like.

  The third quantum dots 16 and the fourth quantum dots 17 are formed on the substrate 11 so as to have a positional relationship such that they are coupled to each other coherently by near-field light. At this time, the third quantum dots 16 and the fourth quantum dots 17 are formed on the substrate 11 so as to be symmetrical via the first quantum dots 12. At this time, an isosceles triangle may be formed between the third quantum dot 16 and the fourth quantum dot 17 with the first quantum dot 12 as a vertex, as shown in FIG.

  Next, the operation of the arithmetic circuit 2 to which the present invention is applied will be described. The arithmetic circuit 2 operates as a product-sum arithmetic element based on a specific photophysical phenomenon that occurs between the quantum dots 12, 13, 16, and 17 formed on the substrate 11. In such a case, an AND operation element functions between the first quantum dot 12 and the quantum dots 16, 17, and a sum operation element between the first quantum dot 12 and the second quantum dot 13. Will function as.

FIG. 9 shows an energy diagram in the case where the material system constituting each quantum dot 12, 16, 17 is CuCl. Based on the above equation (1), E (n _x , n _y , n _z ) of each quantum dot 12, 16, 17 is calculated. At this time, when the side length ratio between the third quantum dot 16 and the fourth quantum dot 17 is about 1: 1, the quantum level in the third quantum dot 16 is (1, 1, 1). E (111) at the time is equal to E (111) when the quantum level in the fourth quantum dot 17 is (1,1,1). That is, the excitation level of the exciton resonates between the quantum level (1, 1, 1) of the third quantum dot 16 and the quantum level (1, 1, 1) of the fourth quantum dot 17. There is a relationship. In fact, in order to cause resonance between them, light having a wavelength corresponding to the quantum level (1, 1, 1) in the first quantum dot 16 is supplied as the signal light SA, or the fourth It is necessary to supply light having a wavelength corresponding to the quantum level (1, 1, 1) in the quantum dot 17 as the signal light SA2.

  When such resonance occurs, excitons existing in the quantum level (1, 1, 1) present in the third quantum dot 16 are converted into the quantum level (1, 1, 1) in the fourth quantum dot 17. The exciton existing at the quantum level (1, 1, 1) of the fourth quantum dot 17 moves to the quantum level (1, 1, 1) of the third quantum dot 16. The excitons are coherently coupled between the quantum dots 16 and 17, and an apparently single excitation mode is formed.

  In other words, the arithmetic circuit 2 is provided with quantum dots 16 and 17 having substantially the same shape and size, each having a side length ratio of 1: 1, on the substrate 11 so that the state density functions become substantially equal. By creating a resonance effect between them, one excitation mode can be formed between the quantum levels (1, 1, 1) of each other.

In FIG. 9, when focusing on the first quantum dot 12a and the third quantum dot 16a and the fourth quantum dot 17a formed in the vicinity thereof, the signal light SA1 is supplied to the third quantum dot 16a. In addition, as a result of the signal light SA2 being supplied to the fourth quantum dot 17a, two excitons are excited from the ground level to the excitation energy level P ₂ having an energy difference of 2h ₁ (double exciton state). Is shown. When this state is the initial state, one exciton is injected into the first quantum dot 12a. As a result, one exciton exists between the third quantum dot 16a and the fourth quantum dot 17a. The case where one exciton exists in the first quantum dot 12a is called a final state.

In this final state, only one exciton is excited between the third quantum dot 16a and the fourth quantum dot 17a constituting the arithmetic unit 10, and therefore, according to the near-field optical coupling strength U. Energy is separated. When the separated energy has the energy level S ₂ as the high level side, the energy level A ₂ as the low level side, and the first quantum dot 12 a has an energy difference of h ₂ from the ground level. In addition, the energy difference between the energy level S ₂ and the ground level is h ₁ + h ₂ + U.

Here, since resonance occurs when the excitation energy level P ₂ in the initial state becomes equal to the energy level S ₂ in the final state, the resonance condition is represented by h ₁ + h ₂ + U = 2h ₁ . Here, when the resonance conditional expression is arranged, Δh = h ₂ −h ₁ = −U, and the quantum level (energy level S ₂ ) in the first quantum dot 12a is more in the near-field optical coupling than h ₁ . The resonance effect can be obtained by adjusting the intensity U to be lower.

Excitons that are injected into the quantum level S ₂ in the first quantum dot 12a is released to a lower energy level. By the way, if the lower energy level is in resonance with the one quantum level in the second quantum dot 13 described above, the emitted exciton is used as the signal light A to the second quantum dot 13. It can be moved.

The energy level A ₂ and the excitation energy level P ₂ are forbidden by the mechanism described above and thus do not resonate.

  That is, in the arithmetic circuit 1 to which the present invention is applied, when signal light is supplied to both the third quantum dot 16 or the fourth quantum dot 17, in other words, the H level signal is supplied to the quantum dots 16 and 17, respectively. Is supplied, excitons are released by the resonance effect described above, and move to the second quantum dots 13. This corresponds to an H level input signal being supplied to the first quantum dot 12.

  On the other hand, when no signal light is supplied to either the third quantum dot 16 or the fourth quantum dot 17, or any of the third quantum dot 16 or the fourth quantum dot 17 When signal light is supplied to only one side, the above-described resonance does not occur, and exciton emission does not occur, so that it does not move to the second quantum dot 13. This corresponds to an L level input signal being supplied to the fourth quantum dot 17.

  For this reason, the arithmetic unit 10 and the first quantum dot 12 in the arithmetic circuit 2 to which the present invention is applied receive (H, H) signals from the third quantum dot 16 and the fourth quantum dot 17, respectively. When supplied, an H level signal is output, and when an (H, L), (L, H), (L, L) signal is supplied, an L level signal is output. It acts as a so-called AND operation element.

  Note that the arithmetic circuit 2 to which the present invention is applied is a so-called symmetric system in which the third quantum dot 16 and the fourth quantum dot 17 are symmetrical to each other via the first quantum dot 12. It is assumed that. When the population in the case of one exciton state is calculated in such a symmetrical system, the first quantum dot 12 is positive with respect to each quantum dot 16 and 17 as shown in FIG. It can be seen that, when gradation is performed (Δh = + U), resonant energy transfer from the calculation unit 10 to the first quantum dots 12 occurs. Incidentally, the population in such a case will rise to a probability of 0.5.

  Further, as shown in FIG. 10B, the population in the biexciton state is obtained when the first quantum dot 12 is negatively gradation with respect to the quantum dots 16 and 17 (Δh = It can be seen that resonance energy transfer from the arithmetic unit 10 to the first quantum dot 12 occurs in -U). The population in such a case will rise to a probability of around 1.0.

In the above-described embodiment, the description will be made by taking as an example a case where the quantum level (excitation energy level S ₂ ) in the first quantum dot 12 is adjusted to be lower than h ₁ by the intensity U of near-field light coupling. However, the present invention is not limited to this. If the excitation energy level is lower than h _1, it is possible to obtain substantially the same effect as described above.

  In addition, the two quantum dots 16 and 17 and one first quantum dot 12 constituting the arithmetic unit 10 acting as the AND arithmetic element are used as a unit and are discretely formed around the second quantum dot 13. By doing so, it is possible to generate output light as an output signal obtained by counting the number of H level signal lights output from the AND operation element.

  For example, when the signal levels of the signal lights SA1 and SA2 are (L, L) and the signal levels of the signal lights SB1 and SB2 are (H, L), excitation transmitted to the second quantum dot 13 The amount of children is also the smallest, and the light intensity L0 of the output light based on these emissions is minimized.

  On the other hand, when the signal levels of the signal lights SA1 and SA2 are (H, H) and the signal levels of the signal lights SB1 and SB2 are (H, L), they are transmitted to the second quantum dot 13. The excitons that are generated consist only of the first quantum dots 12a, and the light intensity L1 of the output light based on these emissions also increases.

  On the other hand, when the signal levels of the signal lights SA1 and SA2 are (H, H) and the signal levels of the signal lights SB1 and SB2 are (H, H), they are transmitted to the second quantum dot 13. The excitons to be generated are the first quantum dots 12a and the first quantum dots 12b, and the light intensity L2 of the output light based on these emissions is nearly twice as large as L1.

  That is, the result of summing up the number of logical sums of signal light only by identifying the light intensity of output light by investigating the relationship of the number of signals supplied to the light intensity of these output lights in advance. Can be obtained. In particular, in the arithmetic circuit 2 to which the present invention is applied, a product-sum operation can be performed between nanometer-sized quantum dots without being governed by the diffraction limit of light. Can also be realized in nano dimensions.

Incidentally, in the arithmetic circuit 2 to which the present invention is applied, as shown in FIG. 11, with respect to input signals a _{1 to} a _N represented as N-bit information, reference signals b as N elements prepared in advance. ₁ ~b _N can also perform so-called matching operation to determine whether the same. In such a case, as shown in the following equation (3), the input signal is multiplied by the reference signal for each bit to obtain the total sum Y0 of the obtained products.

  A matching operation is realized by identifying how similar the calculated product sum Y0 is to the element indicated as the reference signal.

  In the following description of the matching calculation example, two internal variables are defined. The logic 1 when the signal level is H is expressed as (10), and the logic 0 when the signal level is L is expressed as (01). At this time, the 4-bit input signal (1010) is expressed as a = (10011001) as shown in FIG. On the other hand, when the reference signal is (1010), b = (10011001) by the same information expression.

  Here, when a product operation is performed for each bit, the result is a · b = (10011001) as shown in FIG. For the obtained 8 bits, 4 is obtained when the sum operation is performed. Since this is equal to the number of bits of the input signal, it can be determined that the input signal matches the reference signal.

  Similarly, when the 4-bit input signal is (1110), the above information expression represents a = (10101001) as shown in FIG. When this input signal is multiplied by the reference signal b, a · b = (10001001) is obtained as shown in FIG. For the 8 bits obtained, 3 is obtained when the sum operation is performed. Since this differs from the number of bits of the input signal, it can be determined that the input signal and the reference signal do not match.

  Further, this reference signal may be expressed as table data including a so-called arbitrary bit (Don'Care), and this may be matched with the input signal. For example, as shown in FIG. 14, when an arbitrary bit is expressed by (11), the product operation with this arbitrary bit is 1 for both logic 1 and logic 0.

  For example, as shown in FIG. 14, table data (1 * 10) including an arbitrary bit * is represented by b = (10111001) in the above information expression, but in contrast to this, a 4-bit input signal (1110) Multiplying (represented as a = (10101001) by the above information representation) yields a · b = (10101001). When the sum operation is performed for the obtained 8 bits, 4 is obtained.

  Similarly, when a 4-bit input signal (1010) (expressed as a = (10011001) by the above information expression) is multiplied, a · b = (10011001) is obtained. When the sum operation is performed for the obtained 8 bits, 4 is obtained.

  That is, the result of the product-sum operation is 4 regardless of whether the 4-bit input signal is (1110) or (1010).

  Providing such an arbitrary bit plays an important role in Internet routing in which Longest Prefix Match determination is required.

  When the above-described matching operation is performed by the arithmetic circuit 2 to which the present invention is applied, the third quantum dot 16 and the fourth quantum that function as AND operation elements so that a 4-bit input signal can be supplied. The dots 17 and the first quantum dots 12 are taken as one unit, and four units are formed on the substrate 11. Then, the signal level is switched to H or L according to the logic 1 and logic 0 of the 4-bit input signal, and this is supplied to the third quantum dot 16 of the arithmetic unit 10. Further, the signal level is switched according to logic 1, logic 0, or an arbitrary bit for a preset 4-bit reference signal, and this is supplied to the fourth quantum dot 17 of the arithmetic unit. As a result, it is possible to obtain the sum of logical products calculated based on Expression (3) from the light intensity of the output light as the output signal from the second quantum dot 13.

Incidentally, in the arithmetic circuit 2 to which the present invention is applied, when the input signal a represented as N-bit information is expressed by 2 bits, for example, as shown in FIG. 15, the i-th bit is (a _{i, 0} , A _{i, 1} ), and the i-th bit of the reference signal corresponding thereto is also expressed as (b _{i, 0} , b _{i, 1} ). In the AND operation element at this time, a _{i, 0} · b _{i, 0} and a _{i, 1} · b _{i, 1} are calculated, and finally the product-sum operation as shown in the following equation (4) The result Z0 is obtained as an output signal.

In such a case, since reference signals b _{1 to} b _N as N elements prepared in advance are supplied to an N-bit input signal, a total of 2N signals are supplied to the arithmetic unit 10. However, the present invention is not limited to this case. For example, as shown in FIG. 16, the digitized input signal is divided into a plurality of parts, and the divided input signals are supplied to the arithmetic unit 10 at time intervals. You may do it.

In such a case, first, the bit of the input signal is divided into two, a _{1,0 to} a _{N, 0} and a _{1,1 to} a _{N, 1} . Among the divided input signals, a _{1,0 to} a _{N, 0} are supplied to the arithmetic unit 10 at time t = 0, and the reference signals b _{1,0 to} b _{N, 0 corresponding} thereto are simultaneously supplied. As a result, for the output signal, the product of each bit and the sum thereof are obtained based on the equation (5).

Next, at time t = Δt, among the divided input signals, a _{1,1 to} a _{N, 1} are supplied to the arithmetic unit 10 and the reference signals b _{1,1 to} b _{N, 1} are simultaneously supplied. . As a result, as for the output signal, the product of each bit and the sum thereof are obtained based on Equation (6).

  Finally, by adding these output signals, the product-sum operation result of the reference signal for all bits of the input signal can be obtained. This product-sum operation result can be defined by the above-described equation (4).

  By supplying the divided N-bit input signals to the arithmetic unit 10 at time intervals, it is possible to perform the product-sum operation by simply inputting N signals.

  The arithmetic circuit 2 to which the present invention is applied can also be applied to a so-called contendable memory as shown in FIG.

For example, when the input information A is input into the system, it is identified which of the data B _{1 to} B _M the input information A is most similar to. Then, by returning the addresses of the most similar data B _{1 to} B _M , this can be used as a matching address for the input information A.

  Further, the arithmetic circuit 2 to which the present invention is applied can be further applied to a matching mechanism 3 as shown in FIG.

  In this matching mechanism 3, when a code string of the optical code division multiplexed signal D1 represented by “p” indicating the signal level H and “0” indicating the signal level L is input. This is replicated according to the number of reference signals J to be referred to at the time of matching. If there are N reference signals, N copies are made for this signal D1.

Next, the arithmetic circuit 2 described above, and duplicated signals D1, performs product-sum operation between the respective reference signals J ₁ through J _N. In such a case, the matching process is performed by supplying each code constituting the signal D1 and the code of the reference signal corresponding thereto to the quantum dots 16 and 17. Incidentally, since one arithmetic circuit 2 is required for the product-sum operation between one reference signal J and the signal D1, when there are N reference signals, the arithmetic circuit 2 is also used. N are required. In the matching mechanism 3, the reference signal J having the highest similarity to the signal D1 can be identified from the obtained product-sum operation result, and the processing is ended by outputting the reference signal J through a display surface (not shown). To do.

It is a figure which shows the structure of the arithmetic circuit by the quantum dot to which this invention is applied. It is a figure for demonstrating about the energy level of the quantum dot in the arithmetic circuit by the quantum dot to which this invention is applied. It is a figure shown about the simulation result of the time change of population. It is a figure which shows the experimental result about the light intensity change of the output light from a 2nd quantum dot. It is another figure which shows the experimental result about the light intensity change of the output light from a 2nd quantum dot. It is a figure which shows the result of having evaluated the mode of condensing from the 1st quantum dot to the 2nd quantum dot in the space region. It is a figure which shows the structure of the arithmetic circuit by the quantum dot which can perform a product-sum operation process. It is a figure for demonstrating about the positional relationship of the quantum dot formed on the board | substrate in an arithmetic circuit. It is a figure which shows the dynamics in a symmetrical system. It is a figure which shows the time-dependent change of the population in each exciton state. It is a diagram illustrating an example of performing a matching operation on the input signal a ₁ ~a _N expressed as information of N bits. It is a figure which shows the example of a matching calculation in case a is set to (10011001). It is a figure which shows the example of a matching calculation in case a is set to (10101001). It is a figure which shows the example of a matching calculation in the case of providing an arbitrary bit. The input signal a represented as N-bit information is a diagram for explaining the case where it is expressed by 2 bits. It is a figure for demonstrating about the example which divides | segments the digitized input signal into several and supplies the divided | segmented input signal to a calculating part at a time interval mutually. It is a figure for demonstrating about the example which applies the arithmetic circuit to which this invention is applied to what is called a contendable dressable memory. It is a figure for demonstrating about the example which applies the arithmetic circuit to which this invention is applied to a matching mechanism.

Explanation of symbols

1, 2 arithmetic circuit, 3 matching mechanism, 11 substrate, 12 1st quantum dot, 13 2nd quantum dot, 16 3rd quantum dot, 17 4th quantum dot, 20 quantum dot group

Claims (8)

In an arithmetic circuit that generates an output signal represented as the sum of a plurality of input signals,
A dielectric substrate;
A plurality of quantum beams of different sizes corresponding to the input signal are supplied so that excitons are excited to energy levels according to the frequency of the supplied signal light. An input-side quantum dot group in which dots are formed on the substrate;
According to resonance with each energy level, each quantum dot constituting the input-side quantum dot group has a resonance energy level in which excitons are injected through the substrate, and is emitted from the resonance energy level. And an output-side quantum dot that generates, as the output signal, output light having a wavelength longer than that of each of the signal lights in accordance with the received energy. The arithmetic circuit according to claim 1, wherein the output-side quantum dots are formed in a larger volume than the quantum dots constituting the input-side quantum dot group. In an arithmetic circuit that generates an output signal expressed as a sum of products of a plurality of input signals,
A dielectric substrate;
The first signal light and the second signal light corresponding to the input signal are respectively supplied, and a first energy level having a first energy level in which excitons are excited in accordance with the supplied first signal light. And a second energy level that is the same level as the first energy level, and excitons are excited at the second energy level in response to the supplied second signal light. And a third quantum dot having a third energy level lower than the first and second energy levels, through the third quantum dot. A plurality of input-side quantum dot groups formed on the substrate such that the first quantum dots and the second quantum dots are symmetrical to each other;
In accordance with the resonance with the third energy level in the third quantum dot, the input quantum dot group has a resonance energy level from which excitons are injected, and is emitted from the resonance energy level. An arithmetic circuit comprising: an output-side quantum dot that generates output light as the output signal according to energy. The arithmetic circuit according to claim 3, wherein the output-side quantum dots are formed in a larger volume than a third quantum dot that constitutes the input-side quantum dot group. In an arithmetic circuit that generates an output signal represented as a product sum of a reference signal with respect to a digitized input signal,
A dielectric substrate;
The first signal light corresponding to each bit of the input signal is supplied, and the second signal light corresponding to each bit of the reference signal to be integrated with each bit is supplied and supplied. A first quantum dot having a first energy level in which excitons are excited in response to the first signal light, and a second energy level having the same level as the first energy level. A second quantum dot in which excitons are excited to the second energy level according to the supplied second signal light, and a level lower than the first and second energy levels. A third quantum dot having a third energy level is formed on the substrate so that the first quantum dot and the second quantum dot are symmetrical to each other through the third quantum dot. With multiple input quantum dot groups ,
In accordance with the resonance with the third energy level in the third quantum dot, the input quantum dot group has a resonance energy level from which excitons are injected, and is emitted from the resonance energy level. An arithmetic circuit comprising: an output-side quantum dot that generates output light as the output signal according to energy. The arithmetic circuit according to claim 5, wherein the output-side quantum dots are formed with a larger volume than a third quantum dot that constitutes the input-side quantum dot group. Dividing means for dividing the digitized input signal into a plurality of parts and supplying the divided input signals to the input-side quantum dot group at time intervals from each other;
6. The arithmetic circuit according to claim 5, further comprising adding means for adding output signals output from the output side quantum dots at the time interval. In an arithmetic circuit for calculating the similarity of each digitized input signal to each reference signal,
A plurality of arithmetic processing means to which the above input signal and different reference signals are respectively supplied;
A similarity identifying means for identifying the similarity based on an output signal output from each arithmetic processing means;
Each arithmetic processing means is
A dielectric substrate;
The first signal light corresponding to each bit of the input signal is supplied, and the second signal light corresponding to each bit of the reference signal to be integrated with each bit is supplied and supplied. A first quantum dot having a first energy level in which excitons are excited in response to the first signal light, and a second energy level having the same level as the first energy level. A second quantum dot in which excitons are excited to the second energy level according to the supplied second signal light, and a level lower than the first and second energy levels. A third quantum dot having a third energy level is formed on the substrate so that the first quantum dot and the second quantum dot are symmetrical to each other through the third quantum dot. With multiple input quantum dot groups ,
In accordance with the resonance with the third energy level in the third quantum dot, the input quantum dot group has a resonance energy level from which excitons are injected, and is emitted from the resonance energy level. An arithmetic circuit comprising: an output-side quantum dot that generates output light as the output signal according to energy.

Images