Demonstration of quantum permutation parity determine algorithm in a superconducting qutrit*

Project supported by the National Key Basic Research and Development Program of China (Grant No. 2016YFA0301802) and the National Natural Science Foundation of China (Grant Nos. 11504165, 11474152, and 61521001).

Dai Kunzhe1, Zhao Peng1, Li Mengmeng1, Tan Xinsheng1, †, Yu Haifeng1, 2, †, Yu Yang1, 2
National Laboratory of Solid State Microstructures, School of Physics, Nanjing University, Nanjing 210093, China
Synergetic Innovation Center of Quantum Information & Quantum Physics, University of Science and Technology of China, Hefei 230026, China

 

† Corresponding author. E-mail: txs.nju@gmail.com hfyu@nju.edu.cn

Project supported by the National Key Basic Research and Development Program of China (Grant No. 2016YFA0301802) and the National Natural Science Foundation of China (Grant Nos. 11504165, 11474152, and 61521001).

Abstract

A quantum algorithm provides a new way in solving certain computing problems and usually faster than classical algorithms. Here we report an implementation of a quantum algorithm to determine the parity of permutation in a single three-dimensional (3D) superconducting transmon qutrit system. The experiment shows the capacity to speed up in a qutrit, which can also be extended to a multi-level system for solving high-dimensional permutation parity determination problem.

1. Introduction

Quantum computation has been fascinating as the great potential power in solving the computational tasks more efficiently than the classical computer.[1] Over the past few decades, lots of quantum algorithms[2,3] have been proposed and provide a new vision in solving certain problems. Most of the quantum algorithms can realize quantum intrinsic parallel computation in a system with at least two entangled qubits, e.g., the Deutsch-Josza algorithm,[4] Shor’s factoring algorithm,[5] Grover’s searching algorithm,[6] and so on.[710] Generally, superposition and entanglement are considered as the main sources of acceleration of quantum computation. Mark Howard et al.[11] proposed that quantum contextuality[12] also plays an important role in determining the power of quantum computing; although the origin of the speed-up in quantum algorithms has been studied a lot, it is still not completely clear.[13] Recently, an interesting and simple quantum algorithm[14,15] is proposed to realize computational speed-up and the implementation requires only a single qutrit. This quantum algorithm can determine the parity of permutation twice as fast as the classical algorithm. It has been experimentally demonstrated in different types of quantum systems, including the nuclear magnetic resonance systems[14,15] and optical systems.[16,17] As the single qutrit is the smallest system[18] where the quantum contextuality can be observed, these experiments support the points that the entanglement is not necessary in all quantum speed-up, and the contextuality may be the origin of the speedup in this algorithm.

As one of the most promising candidates for building a quantum computer, superconducting quantum circuits can also be used to realize this quantum algorithm. In superconducting quantum circuits, people usually use the lowest two levels of the transmon to serve as a qubit for designing quantum computation algorithms. However, it is well known that a single superconducting transmon actually possesses multiple energy levels and can be treated as an artificial atom. A multi-level system, which corresponds to a large Hilbert space, can not only be used to design a Hamiltonian for quantum simulations,[19,20] but also store or encode more quantum information,[2125] so that it can perform a much more complicated quantum algorithm than a qubit. In this paper, we demonstrate the permutation algorithm in a superconducting 3D transmon qutrit, which is the simplest case to show the computational speed-up. Our scheme holds the promise to extend the permutation algorithm to N dimensions with N levels or N qubits, providing a platform for studying the origin of speed-up in a superconducting quantum circuit system.

2. Quantum permutation algorithm

Considering there is a black box to realize cyclic permutations, in the three objects case, the six different permutations can be defined in the set of 1, 2, 3, which are (1, 2, 3), (2, 3, 1), (3, 1, 2), (3, 2, 1), (2, 1, 3), and (1, 3, 2). The parity of these permutations can be categorized as even or odd, depending on whether the number of the performed exchange operations would be even or odd. Therefore, the first three are even permutations and the last three are odd permutations. The computational task is to determine the parity of the cyclic permutations realized by the black box after query. In the quantum setting, the input and the output are the qutrit state |m⟩, where m = 0, 1, 2 denote the three energy levels of the qutrit. The permutation can be considered as a function y = f(x) which maps the input x ∈ 0, 1, 2 to the output y ∈ 0, 1, 2. These permutation functions can be realized by the unitary transformation acted on the qutrit. The six permutation functions written in Cauchy’s two-line notation and the corresponding unitary transformations are given in the Eq. (1) below.

In the classical scheme, to determine the parity of permutations requires the query on the black box (function) for at least two different inputs. However, the quantum permutation algorithm has been proposed to solve this problem with only one step of query. In the quantum permutation algorithm, the initial state is prepared in the ground state |0⟩ of the qutrit. By doing a qutrit quantum Fourier transformation (QFT)[1] defined as we transform the qutrit state into the superposition state Then the superposition state is sent into the black box. After the permutation operation (Uk or fk, where k = 1, 2, …, 6 labels the six permutations in Eq. (1)) in the back box, the state can be written as Finally, an inverse QFT is carried out and the final state can be measured. The total flow diagram of the algorithm is shown in Fig. 1. When the parity of the permutation is even (k = 1, 2, 3), we obtain the final state |0⟩ with global phases 0, exp(±2πi/3). For odd permutation (k = 4, 5, 6), the result state is |1⟩ with global phases 0, exp(±2πi/3). Thus, a single step of query is sufficient to determine the parity.

Fig. 1. (color online) The flow diagram of the quantum permutation algorithm. The black box contains an unknown permutation operation given in Eq. (1). Before and after the permutation, there is a QFT and inverse QFT operation which is different from the classical scheme.
3. Experimental details

The experiment system is a superconducting transmon embedded in a 3D aluminum cavity. The main purpose of the cavity in our experiments is to provide a clean electromagnetic environment and to serve as a convenient tool for manipulation and measurement.[19,26] The sample is an Al–AlOx–Al Josephson junction about the size of 150 nm × 150 nm connected to two aluminum pads, which is fabricated on a sapphire substrate. The whole pattern is fabricated with e-beam lithography and double angle e-beam evaporation technique. Here we use the lowest three energy levels of the transmon, namely |0⟩, |1⟩, and |2⟩, as a qutrit system. From spectroscopy measurement, we obtain the transition frequency between |0⟩ and |1⟩ is ω01 = 2π × 7.1572 GHz, and the transition frequency between |1⟩ and |2⟩ is ω12 = 2π × 6.8154 GHz. From these transition frequencies, we can calculate the anharmonicity of the qutrit α = 2π × 341.8 MHz. The relatively large anharmonicity ensures that the manipulation applied on any two level transitions would not cause other spurious excitations. The resonant frequency of the 3D cavity is ωc = 2π × 9.0524 GHz. The coupling strength between the qutrit and the cavity is about 2π × 50 MHz. In our design, the detuning between the qutrit transition frequency and cavity frequency are much larger than the coupling strength between the qutrit and the cavity, thus the system works in the dispersive region.[27] The energy relaxation time T1 of the qutrit is about 10 μs and the decoherence time T2 measured from the Ramsey experiment is about 8 μs. The device is cooled in a cryogen-free dilution refrigerator to a base temperature of about 20 mK. The amplifiers and filters are used in our measurement circuit to increase the signal-to-noise ratio and isolate the device from external noise.

In our experiment setup, two microwave vector signal generators combined with four digital to analog converter (DAC) channels of an Arbitrary Wave Generator (AWG) of the model Tektronix 5014c is used to irradiate the transmon qutrit on-resonance with the transition frequency ω01 and ω12 respectively. The four DAC channels generate square waveforms of different amplitudes and lengths for IQ (In-phase and Quadrature) mixers on the vector signal generators to realize a different XY rotation of the qutrit. The amplitude corresponding to the driving strength is firstly calibrated by the Rabi oscillation experiments. Another microwave source is used to read out the qubit states by measuring cavity transmission. Here we choose the “high power readout” scheme,[19] which provides sufficient noise-to-signal ratio of our data and simplifies our experiment procedures and data analysis. Figure 2 shows our experimental setup for manipulating the 3D transmon qutrit.

Fig. 2. (color online) Schematic diagram of experimental setup for manipulating a 3D superconducting transmon qutrit. The MW01 (MW12) source is used to drive the qutrit on-resonance with the transition frequency ω01 (ω12). The DAC channels 1, 3 (2, 4) are used for manipulations along the x axis (y axis). Both driving microwave pulses and probe microwave pulse are sent into the device through a combiner.
4. Results

Initially, we prepared the qutrit on the ground state |0⟩. The QFT is then implemented by a sequence of six selective rotation pulses , , , , , and as shown in Fig. 3. The selective rotation denotes the rotation between qutrit basis states |m⟩ and |n⟩ along the axes α with the rotation angle θ. This combination has a factor −i different from the QFT described in Eq. (2), which can be seen as a global phase, thus has no impact on the experiment result. The six unitary transformations that compose the permutation operations in the black box are also realized by the combination of different selective rotations. After the black box, an inverse pulse sequence of QFT is applied to realize the inverse QFT. In order to extract the complete information of the result, we employ the state tomography technique[28] to characterize the final state. As the final states are always projected on |0⟩ and |1⟩, being independent on whether the parity of the permutation is odd or even, we can consider the states |0⟩ and |1⟩ as the query register and the state |2⟩ as the auxiliary state. Therefore, we can only perform the two level state tomography on the query register, which simplifies the experiment and thus realize the quantum algorithm with a higher fidelity. Generally, the state of a two level system can be described with the density matrix , where I is the unit matrix, r is the state vector on the Bloch sphere, and is the Pauli vector. We can reconstruct the density matrix completely by obtaining the information of state components along x, y, and z axes. In our experiment, we use the MW01 source in Fig. 2 whose frequency is on-resonance with the transition frequency ω01 to generate the tomography pulse sequence. By measuring populations in the states |0⟩ and |1⟩ after a selective rotation of angle 90° along the x and y axes on the Bloch sphere, we can obtain the x and y components of the final state, respectively. The z component can be directly readout without any additional rotation from the populations in the states |0⟩ and |1⟩.

Fig. 3. (color online) The total pulse sequence of experiment.

Figure 4 shows the results of our implementation of permutation algorithm, with panels (a) and (b) for the permutation U1 and panels (c) and (d) for the permutation U4. Following the principles discussed above, we reconstruct the density matrices (including the real (Re) and imaginary (Im) parts) of the final state. The results agree with the expectation, the U1 operation can be determined as even permutation, while the U4 operation can be determined as odd permutation. The fidelity we obtained is 0.9836 for the even permutation U1 and 0.9641 for the odd permutation U4. The fidelity of the other permutations U2, U3, U5, and U6 are 0.9049, 0.9086, 0.8392, and 0.8534, respectively. The final state resulting from each permutation is plotted on the Bloch sphere, which can be seen in Fig. 5.

Fig. 4. (color online) The state tomography of the final state. Panels (a) and (b) ((c) and (d)) are the real and imaginary parts of the final state under the even (odd) permutation operation U1 (U4).
Fig. 5. (color online) The final state on the Bloch sphere. Panels (a), (b), and (c) correspond to the final state (predicted to be |0⟩) resulting from the even permutation U1, U2, and U3. Panels (d), (e), and (f) correspond to the final state (predicted to be |1⟩) resulting from the odd permutation U4, U5, and U6.
5. Conclusion

In this paper, we propose an experimental realization of the quantum permutation algorithm based on a 3D transmon qutrit and demonstrate the implementation of this algorithm to determine the parity of two permutations. The decoherence and the calibration imperfection are the main limits to the fidelity of our experiment, while the fidelity we obtained is still high enough to determine the parity of the permutations. Employing the third state of the transmon realizes the single step query speed-up in this experiment, which shows the potential power of a multi-level system in processing quantum information. This experiment also provides a possible way for theoretical discussion and exploring the origin of quantum speed-up.

Reference
[1] Nielsen M A Chuang I L 2000 Quantum Computation and Quantum Information Cambridge Cambridge University Press 10.1017/CBO9780511976667
[2] Montanaro A 2016 npj Quantum Information 2 15023
[3] Cleve R Ekert A Macchiavello C Mosca M 1998 Proc. R. Soc. Lond. A 454 339
[4] Deutsch D Jozsa R 1992 Proc. R. Soc. Lond. A 439 553
[5] Shor P W 1997 SIAM J. Comput. 26 1484
[6] Grover L K 1997 Phys. Rev. Lett. 79 325
[7] Farhi E Goldstone J Gutmann S Lapan J Lundgren A Preda D 2001 Science 292 472
[8] Han K H Kim J H 2000 Proceedings of the Congress on Evolutionary Computation July, 2000 1354 1360
[9] Kassal I Jordan S P Love P Mohseni M Aspuruguzik A 2008 Proc. Natl. Acad. Sci. USA 105 18681
[10] Ambainis A 2007 SIAM J. Comput. 37 210
[11] Howard M J Wallman J J Veitch V Emerson J 2014 Nature 510 351
[12] Zhang X Um M Zhang J H An S M Wang Y Deng D L Shen C Duan L M Kim K 2013 Phys. Rev. Lett. 110 070401
[13] Den N Maarten V Phys. Rev. Lett. 2013 110 060504
[14] Dogra S Arvind Dorai K 2014 Phys. Lett. A 378 3452
[15] Gedik Z Silva I A Cakmak B Karpat G Vidoto E L G Soarespinto D O Deazevedo E R Fanchini F F 2015 Sci. Rep. 5 14671
[16] Wang F R Wang Y L Liu R F Chen D X Zhang P Gao H Li F L 2015 Sci. Rep. 5 10995
[17] Zhan X Li J Qin H Bian Z H Xue P 2015 Opt. Express 23 18422
[18] Kochen S Specker E P 1967 J. Math. Mech. 17 59
[19] Tan X S Zhao Y X Liu Q Xue G M Yu H F Wang Z D Yu Y 2017 npj Quantum Materials 2 s41535
[20] Houck A T’ureci H E Koch J 2012 Nat. Phys. 8 292
[21] Kiktenko E O Fedorov A K Manko O V Manko V I 2015 Phys. Rev. A 91 042312
[22] Kiktenko E O Fedorov A K Strakhov A A Manko V I 2015 Phys. Lett. 379 1409
[23] Lanyon B P Barbieri M Almeida M P Jennewein T Ralph T C Resch K J Pryde G J Obrien J L Gilchrist A White A 2009 Nat. Phys. 5 134
[24] Muthukrishnan A Stroud C R 2000 Phys. Rev. A 62 052309
[25] Zhao J Tan X S Lan D Wu H T Xue G M Yu H F Yu Y 2017 Phys. Status Solidi 254 1600640
[26] Rigetti C Gambetta J M Poletto S Plourde B L T Chow J M Córcoles A D Smolin J A Merkel S T Rozen J R Keefe G A Rothwell M B Ketchen M B Steffen M 2012 Phys. Rev. B 86 100506
[27] You J Q Nori F 2003 Phys. Rev. B 68 064509
[28] Bianchetti R Filipp S Baur M Fink J M Lang C Steffen L Boissonneault M Blais A Wallraff A 2010 Phys. Rev. Lett. 105 223601