Neutral atom quantum bit control system and method based on real-time calculation

Abstract

The invention relates to a neutral atom quantum bit control system and method based on real-time calculation. The neutral atom quantum bit control system comprises an image acquisition device, a control device, a digital-to-analog conversion device, an analog front-end device and an acousto-optic control device, the image acquisition device is used for acquiring a fluorescence image of an atomic array and sending the fluorescence image to the control device; the control device performs pipelined real-time calculation according to the fluorescence image of the atom array to obtain an initial atom arrangement condition of each row of the atom array, and generates a digital waveform corresponding to a movement strategy of each step when atoms of each row of the atom array are rearranged according to the initial atom arrangement condition; the digital-to-analog conversion device converts a digital waveform into an analog waveform and then sends the analog waveform to the acousto-optic control device through the analog front-end device; and the acousto-optic control device generates a radio frequency signal based on the analog waveform to control the optical tweezers in the atomic array. The method can be widely applied to the field of digital signal acquisition and processing.

Description

[0001] This application is a divisional application of the Chinese invention patent application with application number 202310452981.2, application date April 25, 2023, and name “A neutral atom quantum bit manipulation system and method based on real-time computing”. Technical Field

[0002] The present invention relates to the field of digital signal acquisition and processing, and in particular to a neutral atom quantum bit manipulation system and method based on real-time calculation. Background Art

[0003] Neutral atom quantum computers based on optical tweezer arrays have become one of the most likely experimental platforms for large-scale quantum computing and quantum simulation. A key step is to use radio frequency signals to drive the acousto-optic refractor to generate diffracted light of different strengths and positions, forming optical tweezers to bind and drag atoms to any specified position, thereby achieving the manipulation of any single atom.

[0004] The existing control system is based on the structure of computers, image acquisition cards and AWG (Arbitrary Waveform Generator). First, the computer needs to receive the data sent by the camera, analyze the atomic position distribution, and use the algorithm to determine how to distribute the initial randomly distributed atoms to the target position, and then generate the corresponding RF signal to drive the movement of the atoms. Among them, the RF signal comes from the AWG, which is generated by pre-calculating the required waveform through a computer, then transmitted and cached in the AWG cache, and then the waveform output is controlled by control instructions.

[0005] At present, in order to meet the higher driving frequency of acousto-optic crystals, AWG often needs to achieve a sampling rate of more than GHz; in addition, if multiple atoms need to be controlled simultaneously, a superposition signal output of multiple frequencies is required. The high amount of computation and data brought about by this limits the ability of AWG to manipulate atoms in neutral atom quantum computers in more complex and lower latency demand scenarios. In addition, since the information acquisition module, algorithm execution module, waveform control module, etc. are independent of each other, the total time consumption is the sum of the time consumption of each module plus the internal communication delay, and the total time consumption is relatively large. The longer the time consumption, the greater the probability of atomic collision or decoherence, which will limit the number of quantum bits that can be achieved and the fidelity of quantum operations. Summary of the invention

[0006] In response to the above problems, the purpose of the present invention is to provide a highly integrated, low-latency neutral atom quantum bit manipulation system and method based on real-time computing, which can effectively reduce latency and total time consumption by rearranging atoms in parallel.

[0007] To achieve the above object, the present invention adopts the following technical solutions:

[0008] In a first aspect, the present invention provides a neutral atom quantum bit manipulation system based on real-time computing, comprising: an image acquisition device, a control device, a digital-to-analog conversion device, an analog front-end device, and an acousto-optic control device;

[0009] The image acquisition device acquires the fluorescence image of the atomic array and sends it to the control device;

[0010] The control device performs pipeline real-time calculations based on the fluorescent image of the atomic array to obtain the initial atomic arrangement of the atomic array, and calculates the movement strategy of each step when rearranging the atoms in each row of the atomic array based on the initial atomic arrangement, generates a digital waveform corresponding to each step of the movement strategy and sends it to the digital-to-analog conversion device;

[0011] The digital-to-analog conversion device converts the digital waveform into an analog waveform and sends it to the sound and light control device via the analog front-end device;

[0012] The acousto-optic control device generates a radio frequency signal based on an analog waveform to control the optical tweezers in the atomic array, thereby controlling each atom in the atomic array.

[0013] Furthermore, the control device is implemented based on an FPGA chip.

[0014] Furthermore, the control device is provided with an image decoding module, a processing module and a digital waveform generating module;

[0015] The image decoding module is used to decode the fluorescent image row by row to obtain the initial atomic arrangement of each row in the atomic array;

[0016] The processing module is used to calculate the movement strategy of each step based on the initial atomic arrangement of each row in the atomic array and the target atomic structure;

[0017] The digital waveform generation module is used to generate a corresponding digital waveform according to the movement strategy of each step, and send a synchronization signal to the analog front-end device at the same time.

[0018] Further, the image decoding module includes a gray value calculation submodule and an atomic arrangement calculation submodule;

[0019] The gray value calculation submodule is used to translate the fluorescent image into gray value information of each pixel grid point row by row;

[0020] The atomic arrangement calculation submodule is used to translate the gray value information of each pixel grid point into whether each grid point in the optical tweezers array is occupied by an atom, and finally obtain the initial atomic arrangement of each row in the atomic array.

[0021] Furthermore, the processing module includes an initial parameter acquisition submodule, a single-frequency waveform sequence calculation submodule, a multiplier calculation submodule, an addition tree calculation submodule and an instruction output submodule. The initial parameter acquisition submodule is used to obtain the frequency and initial phase of atoms in each row of the atomic array at each moment according to the initial atomic arrangement of each row in the atomic array; the single-frequency waveform sequence calculation submodule is used to generate a single-frequency waveform sequence generation instruction for each step according to the frequency and initial phase of atoms in each row of the atomic array at each moment; the multiplier submodule is used to generate an amplitude adjustment instruction for each step according to the amplitude of the single-frequency waveform sequence at each moment; the addition tree module is used to generate a multi-frequency superposition instruction for each step according to all adjusted single-frequency waveform sequences; and the instruction output module is used to output all instructions for each step to the digital waveform generation module.

[0022] Furthermore, the digital waveform generation module includes a FIFO module, a control and synchronization module, a K X - Mixing module, K Y - Mixing module, multiplexer and FPGA-DAC communication module;

[0023] The FIFO module is used to "cache" the instructions input by the processing module and read them out one by one and send them to the K X - Mixing module and K Y - Mixing module;

[0024] The K X - Mixing module and K Y -The frequency mixing module is used to generate a single-frequency waveform sequence, adjust the amplitude, and perform multi-frequency superposition according to each instruction read out, thereby generating two sets of mixing signals;

[0025] The multiplexer is used to control the corresponding relationship between the two groups of mixing signals and the actual waveform output, and the obtained digital waveform sequence is sent to the digital-to-analog conversion device via the FPGA-DAC communication module;

[0026] The control and synchronization module is used to X Mixing module, K Y The state of the mixer module is monitored. When the current instruction is executed, the next instruction will be read from the FIFO module, otherwise it will continue to wait. At the same time, the control and synchronization module will send a synchronization signal to the analog front-end device to synchronize each operation step.

[0027] Furthermore, the K X - Mixing module and K Y -The two sets of mixing signals output by the mixing module are used to control the movement of the optical tweezers in two directions in the 2D plane, expressed as:

[0028]

[0029] Among them, SX n and SY n represents the amplitude sequence of two waveforms, representing two moving dimensions, n represents the nth clock of the generated waveform sequence, A i,n represents the amplitude of the nth clock of the ith waveform, f i,n represents the "frequency" of the nth clock cycle of the ith waveform, represents the initial phase of the i-th waveform; K X , K Y are the single frequency numbers of the two waveform sequences respectively.

[0030] Furthermore, the analog front-end device includes a power amplifier and a digital step attenuator, which are used to dynamically adjust the radio frequency power input to the acousto-optic control device.

[0031] In a second aspect, the present invention provides a neutral atom quantum bit manipulation method based on real-time computing, comprising the following steps:

[0032] Acquire fluorescence images of the atomic array;

[0033] Using an FPGA control device to perform pipeline real-time calculations based on the acquired fluorescent image of the atomic array, the initial atomic arrangement of the atomic array is obtained, and the movement strategy of each step when rearranging the atoms in each row or column of the atomic array is calculated based on the initial atomic arrangement, and a digital waveform corresponding to each step of the movement strategy is generated;

[0034] The obtained digital waveform is converted from digital to analog, and a radio frequency signal is generated based on the analog waveform to control the optical tweezers in the atomic array.

[0035] Furthermore, the FPGA control device performs pipeline real-time calculations based on the acquired fluorescence image of the atomic array to obtain the initial atomic arrangement of the atomic array, and generates a digital waveform corresponding to each step of the movement strategy when rearranging the atoms in each row or column of the atomic array according to the initial atomic arrangement, including:

[0036] Decode the fluorescence image to obtain the arrangement of the initial atoms in any row of the atomic array;

[0037] Based on the arrangement of the initial atoms in the atomic array and the preset final target structure, the movement strategy for each step is calculated;

[0038] According to the movement strategy of each step, a digital waveform is generated to rearrange each atom in the atomic array;

[0039] Repeat the above steps, perform pipeline calculations, and obtain the movement strategy for each step.

[0040] The present invention adopts the above technical solution, which has the following advantages:

[0041] 1. Since all logic modules are integrated into a single FPGA chip, the system has a high degree of integration compared to the previous system built with general equipment, which greatly reduces the communication delay between modules (hundreds of nanoseconds). Traditional structures often require a delay of more than milliseconds.

[0042] 2. In the present invention, since there is a deterministic delay between all logic modules, parallel operation of modules can be achieved. For example, after the image acquisition module captures the number of atomic occupancy in the first row, the information can be passed to the processing module of the FPGA control device. When the processing module sends instructions to the digital-to-analog conversion module through the digital waveform generation module, it also sends a synchronization signal to the analog front-end device, so that each module remains synchronized. Compared with the traditional architecture, due to the mutual independence of each module, it is impossible to achieve completely strict deterministic timing, so it is necessary to wait for the previous module to complete its work before running the next module. For example, it is necessary to wait for the image acquisition card to capture the photo before starting to use the computer to execute the rearrangement algorithm.

[0043] 3. In the present invention, due to the coordination between various devices, the total time consumption of the system is a small delay plus the longest time for all modules to run, which is a huge improvement compared to the sum of the time consumption of each module in the traditional architecture plus the internal communication delay.

[0044] 4. Since the control device of the present invention is implemented based on FPGA, the digital waveform generation module implemented by its kernel programming enables the system to have extremely high waveform calculation ability, and can calculate and output complex multi-frequency superposition waveforms in real time. The data output rate can reach GSaPS (billion sample points per second) or more, and the delay of the entire output process is at the level of hundreds of nanoseconds. Therefore, by integrating other components, complex modulation operations can be implemented for acousto-optic devices (AOM, AOD), electro-optical devices (EOM), etc. In the traditional architecture, since the AWG itself has no logical operation ability and no real-time generation ability of complex waveforms, the playback of the waveform depends on the computer to complete the real-time calculation and then transmit it to the AWG, or the computer pre-caches it in the AWG. These two methods may be limited by the speed of calculation and transmission or the size of the cache space. Under the requirement of maintaining low latency and fast speed, the ability to manipulate atoms in parallel is very limited.

[0045] 5. The present invention is applied to a neutral atom quantum platform based on an optical tweezers array, which reduces the rearrangement time of the atomic array, thereby reducing the collision loss in the process and increasing the number of achievable quantum bits.

[0046] Therefore, the present invention is expected to become the basic building block of future neutral atom quantum computers based on optical tweezers arrays, and can be widely used in the fields of digital signal acquisition and processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present invention. Throughout the accompanying drawings, the same reference numerals are used to represent the same components. In the accompanying drawings:

[0048] Figure 1a and Figure 1b is a schematic diagram of the rearrangement problem, where Figure 1a is the original arrangement diagram, Figure 1b is the target arrangement diagram;

[0049] Figure 2 is an operation flow chart of a neutral atom quantum bit manipulation system based on real-time calculation in an embodiment of the present invention;

[0050] Figure 3a is a corresponding schematic diagram of implementing the generated waveform and the corresponding movement of the optical tweezers in an embodiment of the present invention;

[0051] Figure 3b is a moving strategy of a single optical tweezer in an embodiment of the present invention;

[0052] Figure 4 is the internal structure of the DWG module in the embodiment of the present invention;

[0053] Figure 5 is a schematic diagram of an addition tree in an embodiment of the present invention;

[0054] Figure 6 It is a timing description of a neutral atom quantum bit manipulation system based on real-time calculation in an embodiment of the present invention;

[0055] Figure 7 It is an example (static) of multi-frequency waveform superposition output generated by real-time calculation of a neutral atom quantum bit manipulation system based on real-time calculation in an embodiment of the present invention;

[0056] Figure 8 It is an example (dynamic) of a multi-frequency waveform superposition output generated by real-time calculation of a neutral atom quantum bit manipulation system based on real-time calculation in an embodiment of the present invention. DETAILED DESCRIPTION

[0057] In order to make the purpose, technical solution and advantages of the embodiment of the present invention clearer, the technical solution of the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings of the embodiment of the present invention. Obviously, the described embodiment is a part of the embodiment of the present invention, not all of the embodiments. Based on the described embodiment of the present invention, all other embodiments obtained by ordinary technicians in this field belong to the scope of protection of the present invention.

[0058] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0059] As shown in Figure 1, as an example, in a specific application, we need to randomly load the initial atoms ( Figure 1a ) are arranged into a compact target grid ( Figure 1b ). Based on this problem, in some embodiments of the present invention, a neutral atom quantum bit manipulation system based on real-time calculation is designed, which is mainly composed of a programmable logic device FPGA (Field Programmable Gate Array), a digital-to-analog conversion device and an analog front end (AFE). In order to obtain the atomic filling situation in the atomic array, the fluorescent image of the atomic array is first read, and the gray value information of each pixel is read and directly transmitted to the FPGA chip. The digital waveform is obtained by real-time calculation by the FPGA and sent to the digital-to-analog conversion device and the analog front end. Finally, the optical tweezers in the atomic array are controlled by the 2D AOD.

[0060] Correspondingly, some other embodiments of the present invention provide a neutral atom quantum bit manipulation method based on real-time computing.

[0061] Example 1

[0062] like Figure 2As shown, this embodiment provides a neutral atom quantum bit manipulation system based on real-time calculation, which includes: an image acquisition device, an FPGA control device, a digital-to-analog conversion device, an analog front-end device, and an acoustic-optical control device. Among them, the image acquisition device is used to collect the fluorescence image of the atomic array and send it to the FPGA control device; the FPGA control device is used to perform pipeline real-time calculations based on the fluorescence image of the atomic array to obtain the initial atomic arrangement of the atomic array, and according to the initial atomic arrangement, the movement strategy of each step when the atoms in each row or column of the atomic array are rearranged, and the digital waveform corresponding to each step of the movement strategy is generated and sent to the digital-to-analog conversion device; the digital-to-analog conversion device converts the digital waveform into an analog waveform, and sends it to the acoustic-optical control device through the analog front-end device; the acoustic-optical control device generates a radio frequency signal based on the analog waveform to control the optical tweezers in the atomic array.

[0063] Preferably, the image acquisition device may be an EMCCD (Electron-Multiplying Charge-coupled Device) camera.

[0064] Preferably, the FPGA control device is provided with an image decoding module (Decoder), a processing module (PS) and a digital waveform generator module (DWG). The image decoding module is used to decode the fluorescent image line by line to obtain the initial atomic arrangement of each row in the atomic array; the processing module is used to calculate the movement strategy of each step based on the initial atomic arrangement of each row in the atomic array and the target atomic structure; the digital waveform generator module is used to generate a corresponding digital waveform according to the movement strategy of each step, and send a synchronization signal to the analog front-end device at the same time.

[0065] Preferably, the image decoding module includes a gray value calculation submodule and an atomic arrangement calculation submodule. The gray value calculation submodule is used to translate the fluorescence image into gray value information of each pixel grid point row by row; wherein, the fluorescence information of a row of atoms may correspond to multiple rows of pixels; the atomic arrangement calculation submodule is used to translate the gray value information of each pixel grid point into whether each grid point in the optical tweezers array is occupied by atoms, and finally obtain the occupation information of each row of atoms, that is, the initial atomic arrangement of each row in the atomic array.

[0066] Preferably, in the processing module, the movement strategy of each step refers to moving one or more atoms from a certain initial position to a specified position at the same time. Due to the limitations of 2D AOD equipment, it is currently only possible to move atoms in a certain row or column at the same time. The emphasis on the simultaneous movement of multiple atoms here is because the present invention can achieve the simultaneous movement of multiple atoms with the lowest delay and the highest degree of freedom. The rearrangement algorithm can wait until the entire image is read and recognized before starting; it can also start to rearrange the atoms in the current row after receiving the atomic information of a certain row. The former is a commonly used solution for the current PC+AWG system; the latter can save the time of rearrangement, but puts higher requirements on the timing of the entire system, which is also the solution supported by the operating system of the present invention.

[0067] In this embodiment, a new row-by-row parallel rearrangement algorithm is proposed by combining the parallel movement capability of atoms and the idea of ​​row-by-row rearrangement, which has excellent effect, low computational complexity and strong scalability. It should be emphasized that in this embodiment, since the processing module itself is a soft-core processor built using the logic resources on the FPGA, computer programming languages ​​such as C language can be executed in it. Compared with hardware description languages ​​represented by Verilog and VHDL, C language is more suitable for describing complex logic algorithms, and is more convenient for debugging and algorithm iteration. The algorithm generated by the processing module is sent to the DWG module in a one-by-one instruction manner.

[0068] Specifically, the processing module includes an initial parameter acquisition submodule, a single-frequency waveform sequence calculation submodule, a multiplier calculation submodule, an addition tree calculation submodule and an instruction output submodule. The initial parameter acquisition submodule is used to obtain the frequency and initial phase of each row of the atoms in the atomic array at each moment according to the initial atomic arrangement of each row in the atomic array; the single-frequency waveform sequence calculation submodule is used to generate a single-frequency waveform sequence generation instruction for each step according to the frequency and initial phase of each row of the atoms in the atomic array at each moment; the multiplier submodule is used to generate an amplitude adjustment instruction for each step according to the amplitude of the single-frequency waveform sequence at each moment; the addition tree calculation submodule is used to generate a multi-frequency superposition instruction for each step according to all adjusted single-frequency waveform sequences; and the instruction output module is used to output all instructions for each step to the digital waveform generation module.

[0069] In order to control the two-dimensional movement of multiple optical tweezers, the multi-frequency superposition instruction output by the addition tree calculation submodule in this embodiment is to simultaneously generate two waveform sequences:

[0070]

[0071] Among them, SX n and SY n represents the amplitude sequence of two waveforms, representing two moving dimensions, n represents the nth clock of the generated waveform sequence, Ai,n represents the amplitude of the nth clock of the ith waveform, f i,n represents the "frequency" of the nth clock cycle of the ith waveform, represents the initial phase of the i-th waveform; K X and K Y are the numbers of single frequencies of the two waveform sequences respectively. In this example, they are 32 and 8 respectively.

[0072] like Figure 3a As shown, due to the Bragg diffraction effect, f i,n In fact, it is proportional to the spatial position of the ith atom at the nth clock. If we control the ith waveform from an initial value f i,0 Move to end value f i,$ , we can control the movement of the corresponding optical tweezers; if we control multiple waveforms to change at the same time, we can control the movement of multiple optical tweezers.

[0073] like Figure 3b As shown, the movement strategy sequence of a single atom is shown, where the intensity corresponds to the parameter A in formula (1) and formula (2). i , the position corresponds to the parameter f in formula (1) and formula (2) i,n , speed corresponds to parameter f i,n The rate of change over time. The whole process is divided into the following five stages:

[0074] a. Generate an initial frequency f i,0 , whose initial phase is

[0075] b. After the preset time t0, the intensity increases from 0 to the maximum value I at a rate k. MAX ;

[0076] c. After waiting for the preset time t1, the speed increases from 0 to the maximum value v at the acceleration a. MAX ;

[0077] d. Keep v MAX uniform motion, then with acceleration -a from v MAX Decelerate to speed 0;

[0078] e. Static at the end frequency f i,t After waiting time t2, the intensity increases from I to MAX Decrease to 0.

[0079] Among them, k,I MAX ,a,v MAX , is a global variable, that is, it is the same for each move and is set before the algorithm starts; and f i,0 ,f i,tThe five variables t0, t1, and t2 are local variables, which need to be set separately for each single frequency each time it moves, so they need to be carried by each instruction.

[0080] In terms of hardware implementation, this embodiment uses a state machine to complete the above steps and output the amplitude A at each moment. i,n and frequency f i,n The state machine uses digital signal synthesis (DDS) to generate a single-frequency signal. It only needs to input the initial phase and the frequency f at each moment i,n , the corresponding single-frequency waveform sequence can be obtained. The biggest advantage of DDS technology is that it can quickly jump the frequency in a phase-continuous manner. The phase continuity makes the output single-frequency signal amplitude continuous, which is crucial for the stability of the optical tweezers intensity and reducing the heating phenomenon when the atoms are bound by the optical tweezers. Apply the single-frequency waveform sequence to a multiplier with a multiplication factor of A. i,n , you can get Figure 3b When the above five steps are completed or waiting to be executed, the state machine is in the "idle" state.

[0081] Preferably, if Figure 4 As shown, the digital waveform generation module includes a FIFO module, a control and synchronization module, a K X - Mixing module, K Y - Mixing module, multiplexer and FPGA-DAC communication module. Among them, the FIFO module is used to "cache" the atomic manipulation instructions input by the processing module and read them out one by one. Each instruction contains K X +K Y Five local variables f of a single frequency signal i,0 ,f i,t ,t0,t1,t2, the corresponding relationship of mux and the synchronization instruction are used to control the analog front-end device; K X - Mixing module and K Y -The mixing module is used to perform mixing according to each instruction read out, that is, the generation of a single-frequency waveform sequence, amplitude adjustment, and multi-frequency superposition to generate two sets of mixing signals; the multiplexer is used to control the corresponding relationship between the two sets of mixing signals and the actual waveform output, and the obtained digital waveform sequence is sent to the digital-to-analog conversion device via the FPGA-DAC communication module; the control and synchronization module is used to control the K X - Mixing module, K Y -The status of the mixer module is monitored. When the status of all modules is "idle", it will start to read the next instruction from the FIFO module, otherwise it will continue to wait. At the same time, the control and synchronization module will send a synchronization signal to the analog front-end device to synchronize each operation step.

[0082] Specifically, Figure 5 As shown, in this embodiment, the digital waveform generation module is designed as a pipelined addition tree module, which can accumulate a preset number of single-frequency input signals (in this embodiment, 2^N single-frequency signal input modules are used, and each single-frequency signal input module generates a single-tone waveform through digital synthesis technology) and output 1 accumulated output result. In this embodiment, a total of 40 single-frequency signal generation modules are instantiated and divided into 2 groups, one of which includes 32 single-frequency signal generation modules and the other includes 8 single-frequency signal generation modules, which are used to generate 2 groups of waveforms. Correspondingly, 2 addition tree modules are instantiated, one for 32 inputs and the other for 8 inputs. Assuming that the bit width of each single-frequency signal is 16 bits, then after 2 N After the input adder, the output mixed signal is (16+N) bits, so a multiplier needs to be designed to normalize the mixed signal and regain a signal with a bit width of 16 bits.

[0083] It should be noted that the two groups of mixing modules instantiated in this embodiment can generate 32 and 8 mixing signals respectively. However, the two mixing modules are not one corresponding to the waveform output of the X dimension and the other corresponding to the waveform output of the Y dimension. If so, in order to enable both the X dimension and the Y dimension to have 32 mixing signal output capabilities, it is necessary to instantiate 2 groups of 32 single-frequency waveform generation modules and 2 groups of 32 input addition trees, which will increase unnecessary logic resource consumption. To this end, we designed a multiplexer module to control the correspondence between the two groups of mixing signals and the actual waveform output. The waveform sequence output by the multiplexer (mux) is pipelined and input into the FPGA-DAC communication module and sent to the DAC.

[0084] We mentioned the pipeline working mode, that is, each small module is responsible for only one sub-function, and generates a processing result in each clock cycle and transmits it to the next module. Therefore, except for the delay when the pipeline starts working, there will be waveform output at every other moment. Figure 6 As shown, the timing of the entire operating system is maximized and time is saved by arranging the various modules of the entire system to work in parallel.

[0085] Preferably, the digital-to-analog conversion device uses a high-speed digital-to-analog converter (DAC, Digital-Analog Converter).

[0086] In this embodiment, the digital-to-analog conversion device used is the DAC39J84 of Texas Instruments, with a 4-channel 16-bit sampling depth and a data input rate of 1.2288×10 9Sample points per second, after 2 times interpolation, the output sampling rate is 2457.6×10 9 Samples per second. Since current FPGAs cannot run at 1.2288×10 9 Hz, so in fact, we adopted 8-way parallel computing, making the clock frequency of DWG running 153.6×10 6 Hz. As an example, Figure 7 This is an example of the multi-frequency waveform superposition output generated by the system in real-time calculation (static). Figure 8 This is an example (dynamic) of the multi-frequency waveform superposition output generated by the system in real-time calculation.

[0087] Preferably, the analog front-end device includes a high power amplifier and a digital step attenuator (DSA), which are used to dynamically adjust the RF power input to the AOD. The control signal of the DSA comes from the control and synchronization module in the DWG, which can be synchronized with each step of the operation, and can maximize the optimization of the output RF power, improve the diffraction efficiency of the AOD and reduce the nonlinear effect.

[0088] Example 2

[0089] Based on the neutral atom qubit manipulation system based on real-time computing provided in Example 1, this embodiment provides a neutral atom qubit manipulation method based on real-time computing, comprising the following steps:

[0090] 1) Acquire a fluorescence image of the atomic array;

[0091] 2) using an FPGA control device to perform pipeline real-time calculations based on the acquired fluorescence image of the atomic array to obtain the initial atomic arrangement of the atomic array, and generating a digital waveform corresponding to each step of the movement strategy when rearranging the atoms in each row or column of the atomic array according to the initial atomic arrangement;

[0092] 3) Performing analog-to-digital conversion on the obtained digital waveform, and generating a radio frequency signal based on the analog waveform to control the optical tweezers in the atomic array.

[0093] Preferably, the above step 2) includes the following steps:

[0094] 2.1) Decode the fluorescence image to obtain the arrangement of the initial atoms in any row of the atomic array;

[0095] 2.2) Based on the arrangement of the initial atoms in the atomic array and the preset final target structure, pipeline calculation is performed to obtain the movement strategy for each step;

[0096] 2.3) generating a digital waveform for rearranging each atom in the atomic array according to the movement strategy of each step;

[0097] 2.4) Repeat steps 2.1) to 2.3) to perform pipeline calculations to obtain the movement strategy for each step.

[0098] Preferably, the above step 2.1) includes the following steps:

[0099] 2.1.1) Translate the fluorescence image into grayscale value information of each pixel grid row by row; where the fluorescence information of a row of atoms may correspond to multiple rows of pixels;

[0100] 2.1.2) The grayscale value information of each pixel grid point is translated into whether each grid point in the optical tweezers array is occupied by an atom, and finally the occupancy information of each row of atoms is obtained, that is, the arrangement of the initial atoms in the atomic array.

[0101] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A neutral atom quantum bit manipulation system based on real-time computing, characterized in that: include: Image acquisition device, control device, digital-to-analog conversion device, analog front-end device and sound and light control device; The image acquisition device acquires the fluorescence image of the atomic array and sends it to the control device; The control device performs pipeline real-time calculations based on the fluorescent image of the atomic array to obtain the initial atomic arrangement of each row of the atomic array, and calculates the movement strategy of each step when rearranging the atoms in each row of the atomic array based on the initial atomic arrangement, generates a digital waveform corresponding to each step of the movement strategy and sends it to the digital-to-analog conversion device; The digital-to-analog conversion device converts the digital waveform into an analog waveform and sends it to the sound and light control device via the analog front-end device; The acousto-optic control device generates a radio frequency signal based on the analog waveform to control the optical tweezers in the atomic array, thereby controlling each atom in the atomic array; The control device is implemented based on an FPGA chip, and includes an image decoding module, a processing module, and a digital waveform generation module; The image decoding module is used to decode the fluorescent image row by row to obtain the initial atomic arrangement of each row in the atomic array; The processing module is used to calculate the movement strategy of each step based on the initial atomic arrangement of each row in the atomic array and the target atomic structure; The digital waveform generation module is used to generate a corresponding digital waveform according to the movement strategy of each step, and send a synchronization signal to the analog front-end device at the same time.

2. A neutral atom quantum bit manipulation system based on real-time computing as claimed in claim 1, characterized in that: The image decoding module includes a gray value calculation submodule and an atomic arrangement calculation submodule; The gray value calculation submodule is used to translate the fluorescent image into gray value information of each pixel grid point row by row; The atomic arrangement calculation submodule is used to translate the gray value information of each pixel grid point into whether each grid point in the optical tweezers array is occupied by an atom, and finally obtain the initial atomic arrangement of each row in the atomic array.

3. A neutral atom quantum bit manipulation system based on real-time computing as claimed in claim 1, characterized in that: The processing module includes an initial parameter acquisition submodule, a single-frequency waveform sequence calculation submodule, a multiplier calculation submodule, an addition tree calculation submodule and an instruction output submodule, wherein the initial parameter acquisition submodule is used to obtain the frequency and initial phase of each row of atoms in the atomic array at each moment according to the initial atomic arrangement of each row in the atomic array; The single-frequency waveform sequence calculation submodule is used to generate a single-frequency waveform sequence generation instruction for each step according to the frequency and initial phase of each row of atoms in the atomic array at each moment; the multiplier submodule is used to generate an amplitude adjustment instruction for each step according to the amplitude of the single-frequency waveform sequence at each moment; the addition tree module is used to generate a multi-frequency superposition instruction for each step according to all adjusted single-frequency waveform sequences; the instruction output module is used to output all instructions for each step to the digital waveform generation module.

4. A neutral atom quantum bit manipulation system based on real-time computing as claimed in claim 1, characterized in that: The digital waveform generation module includes a FIFO module, a control and synchronization module, a K X - Mixing module, K Y - Mixing module, multiplexer and FPGA-DAC communication module; The FIFO module is used to "cache" the instructions input by the processing module and read them out one by one and send them to the K X - Mixing module and K Y - Mixing module; The K X - Mixing module and K Y -The frequency mixing module is used to generate a single-frequency waveform sequence, adjust the amplitude, and perform multi-frequency superposition according to each instruction read out, thereby generating two sets of mixing signals; The multiplexer is used to control the corresponding relationship between the two groups of mixing signals and the actual waveform output, and the obtained digital waveform sequence is sent to the digital-to-analog conversion device via the FPGA-DAC communication module; The control and synchronization module is used to X Mixing module, K Y The state of the mixer module is monitored. When the current instruction is executed, the next instruction will be read from the FIFO module, otherwise it will continue to wait. At the same time, the control and synchronization module will send a synchronization signal to the analog front-end device to synchronize each operation step.

5. A neutral atom quantum bit manipulation system based on real-time computing as claimed in claim 4, characterized in that: The K X - Mixing module and K Y -The two sets of mixing signals output by the mixing module are used to control the movement of the optical tweezers in two directions in the 2D plane, expressed as: Among them, SX n and SY n represents the amplitude sequence of two waveforms, representing two moving dimensions, n represents the nth clock of the generated waveform sequence, A i,n represents the amplitude of the nth clock of the ith waveform, f i,n represents the "frequency" of the nth clock cycle of the ith waveform, represents the initial phase of the i-th waveform; K X , K Y are the single frequency numbers of the two waveform sequences respectively.

6. A neutral atom quantum bit manipulation system based on real-time computing as claimed in claim 1, characterized in that: The analog front-end device includes a power amplifier and a digital step attenuator, which are used to dynamically adjust the radio frequency power input to the acousto-optic control device.

7. A neutral atom quantum bit manipulation method based on real-time computing, characterized in that: The following steps are involved: Acquire fluorescence images of the atomic array; Using an FPGA control device to perform pipeline real-time calculations based on the acquired fluorescent image of the atomic array, the initial atomic arrangement of the atomic array is obtained, and the movement strategy of each step when rearranging the atoms in each row or column of the atomic array is calculated based on the initial atomic arrangement, and a digital waveform corresponding to each step of the movement strategy is generated; The obtained digital waveform is converted into digital-to-analog format, and a radio frequency signal is generated based on the analog waveform to control the optical tweezers in the atomic array; the FPGA control device is used to perform pipeline real-time calculations based on the acquired fluorescence image of the atomic array to obtain the initial atomic arrangement of the atomic array, and the movement strategy of each step when the atoms in each row or column of the atomic array are rearranged according to the initial atomic arrangement, and the digital waveform corresponding to each step of the movement strategy is generated, including: ① Decode the fluorescence image to obtain the arrangement of the initial atoms in any row of the atomic array; ② Based on the arrangement of the initial atoms in the atomic array and the preset final target structure, the movement strategy for each step is calculated; ③ According to the movement strategy of each step, a digital waveform is generated to rearrange the atoms in the atomic array; Repeat the above steps ① to ③ to perform pipeline calculations to obtain the movement strategy of each step and its corresponding digital waveform.

Images