Efficacious qubit mappings for quantum simulations of the 12 C rotational band
Darin C. Mumma 1 , Zhonghao Sun 2 , Alexis Mercenne 1 , Kristina D. Launey 1 , Soorya Rethinasamy 3 4 1 and James A. Sauls 1 4 · 2024
Read Efficacious qubit mappings for quantum simulations of the 12 C rotational band by Darin C. Mumma 1 , Zhonghao Sun 2 , Alexis Mercenne 1 , Kristina D. Launey 1 , Soorya Rethinasamy 3 4 1 and James A. Sauls 1 4 (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-06979.txt.
Introduction
Darin Mumma, Zhonghao Sun, Alexis Mercenne, Kristina Launey, Soorya Rethinasamy, and James Sauls run first quantum simulations of the carbon-12 nucleus in two thousand twenty-four. Solving nuclei from scratch needs computing power that explodes with particle count. Their work targets the low-lying states of carbon-12 with a hybrid quantum method.
Two tricks make the task fit today's noisy machines. A near-perfect symmetry of nuclei shrinks the model space sharply when a symmetry-fit basis is used. A Gray code mapping then lays the problem onto qubits in a lean way, matched to the banded shape of the nuclear force matrix. Even with noise, the runs returned sound bound-state energies.
Catalog record
| CREATOR | Darin C. Mumma 1 , Zhonghao Sun 2 , Alexis Mercenne 1 , Kristina D. Launey 1 , Soorya Rethinasamy 3 4 1 and James A. Sauls 1 4 |
|---|---|
| YEAR | 2024 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 18 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2412-06979
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Efficacious qubit mappings for quantum simulations of the 12 C rotational band Thanks: This work was supported by the U.S. Department of Energy (DE-SC0023694). SR acknowledges support from the School of Electrical and Computer Engineering at Cornell University, from the U.S. National Science Foundation under Grant No. 2315398, and from AFRL under agreement no. FA8750-23-2-0031. We acknowledge discussions with the Quantum Science Center at ORNL. This work benefited from high performance computational resources provided by LSU (www.hpc.lsu.edu), the National Energy Research Scientific Computing Center (NERSC), a U.S. Department of Energy Office of Science User Facility at Lawrence Berkeley National Laboratory operated under Contract No. DE-AC02-05CH11231, as well as the Frontera computing project at the Texas Advanced Computing Center, made possible by National Science Foundation award OAC-1818253. Darin C. Mumma 1 , Zhonghao Sun 2 , Alexis Mercenne 1 , Kristina D. Launey 1 , Soorya Rethinasamy 3 4 1 and James A. Sauls 1 4 Abstract Solving atomic nuclei from first principles places enormous demands on computational resources, which grow exponentially with increasing number of particles and the size of the space they occupy. We present first quantum simulations based on the variational quantum eigensolver for the low-lying structure of the 12 C nucleus that provide acceptable bound-state energies even in the presence of noise. We achieve this by taking advantage of two critical developments. First, we utilize an almost perfect symmetry of atomic nuclei that, in a complete symmetry-adapted basis, drastically reduces the size of the model space. Second, we use the efficacious Gray encoding, for which it has been recently shown that it is resource efficient, especially when coupled with a near band-diagonal structure of the nuclear Hamiltonian. I Introduction Numerical simulations of nuclear structure and reactions are at the frontier of nuclear science research. However, the computational resources typically required scale exponentially with the number of particles and the size of the space they occupy. Quantum computing has the potential to revolutionize the way we compute through the superposition of qubit states [ 1 ] . Earlier studies that aim to solve quantum many-body problems using noisy intermediate-scale quantum (NISQ) computers have been based on various approaches, including the variational quantum eigensolver (VQE) [ 2 , 3 ] , quantum phase estimation [ 4 , 5 ] , and quantum imaginary time evolution [ 6 ] . Among the various quantum algorithms, the VQE approach has been used to solve simplified nuclear science problems [ 7 ] and more realistic problems [ 8 ] for its relatively shallower circuit depths and suitability for near-term quantum computers. Meanwhile, the development of efficient and robust quantum algorithms to solve the nuclear science problem in the Noise Intermediate-Scale Quantum (NISQ) device era [ 9 ] could also help us to understand the entanglement of atomic nuclei, and foster algorithms for classical computing. Fig. 1: Recursive circuit ansatz to generate the superposition of the one-hot basis states used in Fig. ( 3 ). The input to the circuit is |0\rangle^{\bigotimes N} , and \text{params}[i] denote the parameters \theta_{i} with the R_{y}(2\theta_{i})=e^{-i\theta_{i}Y_{i}} rotation gate. Fig. 2: Three-layered circuit ansatz to generate the superposition of the Gray basis states used in Fig. ( 3 ). The input to the circuit is |0\rangle^{\bigotimes N} , and \text{params}[i] denote the parameters \theta_{i} with the R_{y}(2\theta_{i})=e^{-i\theta_{i}Y_{i}} rotation gate. Most nuclei are deformed, and one of the most challenging problems is explaining how fundamental interactions drive the deformation of atomic nuclei. The focus of this research is the structure of 12 C. We perform quantum simulations for the ground state ( 0^{+}_{\rm g.s.} ) and the first 2^{+} excited state of 12 C, which are members of the ground-state rotational band and reflect strong collective correlations (e.g., see [ 10 ] ). Indeed, when starting with single-particle basis states of the harmonic oscillator (HO), the size of the model space required to describe the ground state, e.g., in fourteen HO shells is \sim 10^{11} , which is beyond the memory capabilities of the best of today’s petascale classical computers. Fortunately, when we use a complete symmetry-adapted correlated basis that recognizes an almost perfect symmetry of atomic nuclei [ 11 , 12 ] , the symplectic Sp(3,R) symmetry [ 13 ] , solutions can be achieved in model spaces that are orders of magnitude smaller. As a proof-of-principle, in this study, we provide solutions by adopting the nuclear Hamiltonian of [ 14 ] inspired by a (symplectic) symmetry effective field theory that captures most of the physics, namely, in excess of 85% for each wavefunction (see the contribution of the dominant symmetry-adapted configuration in 0^{+}_{\rm g.s.} and 2^{+} of 12 C from chiral potentials in [ 15 ] ). This Hamiltonian has been shown to successfully describe the elusive Hoyle state in 12 C [ 14 ] . Remarkably, it further reduces the number of basis states to eight states only, enabling early experiments with twelve particles on NISQ processors. Following the first quantum applications to deuteron [ 7 ] , a two-particle system, our quantum simulations use VQE to compute the energy of the twelve-particle nucleus. We use two different qubit encoding schemes, including one-hot and Gray encodings (see, e.g., [ 16 ] ). We explore a recently developed algorithm that can efficiently solve any band-diagonal to full Hamiltonian matrices [ 17 ] . This approach can accommodate realistic nuclear potentials derived from chiral effective field theory, a key ingredient of long-term quantum simulations for ab initio modeling of classically unsolvable atomic nuclei. In particular, we investigate simulations of the 12 C structure with shot noise and noise models from existing quantum devices, and explore the noise-resilient training method [ 18 ] . Our simulations reproduce the theoretical value of the 2^{+} excitation energy in 12 C, even for noisy simulations suitable for the NISQ regime. We study the effectiveness of the one-hot and Gray encodings. We find that the Gray encoding allows an efficient scaling of the Hamiltonian complexity, by using only \log_{2}(N) qubits for N number of basis states, while achieving acceptable outcomes. Indeed, in [ 17 ] , it has been shown that the Gray encoding is more resource efficient not only for tridiagonal Hamiltonians, as suggested earlier in [ 16 ] , but also for band-diagonal Hamiltonians with bandwidth up to N . II Nuclear Hamiltonian and quantum algorithm The nuclear Hamiltonian suitable for NISQ processors emerges from a symmetry effective field theory and consists of a fully microscopic nucleon-nucleon interaction [ 14 ] : \displaystyle H_{\rm SymEFT}= \displaystyle\hskip-17.07182ptH_{0}+\chi\frac{\left(e^{-\gamma Q\cdot Q}-1\right)}{2\gamma} (1) \displaystyle\approx \displaystyle\hskip-17.07182ptH_{0}-\frac{\chi}{2}Q\cdot Q\sum_{k=0}^{\lceil K/2\rceil-1}\frac{(-\gamma)^{k}(Q\cdot Q)^{k}}{(k+1)!}, (2) where H_{0} is the HO Hamiltonian, Q=\sum_{i}\sqrt{16\pi/5}r_{i}^{2}Y_{2}(\hat{\mathbf{r}}_{i}) is the mass quadrupole moment, and {\mathbf{r}}_{i} is the i -th particle coordinate. K<N denotes the number of off-diagonals kept in calculations, with 2K+1 giving the Hamiltonian matrix bandwidth, and \chi and \gamma are parameters of the nuclear potential. The corresponding calculations on classical computers have been earlier used to provide the first no-core shell-model description of the elusive Hoyle state in 12 C that admits cluster substructures [ 19 , 14 ] . Our simulations start from ( 1 ) with \hbar\Omega = 18 MeV, \gamma=1.7\times 10^{-4} , and \chi is defined in [ 14 ] ,