Quantum computing devices have been rapidly developed in the past decade. Tremendous efforts have been devoted to finding quantum advantages for useful but classically intractable problems via current noisy quantum devices without error correction. It is important to know the fundamental limitations of noisy quantum devices with the help of classical computers. For computation with general classical processing, we show that noisy quantum devices with a circuit depth of more than $O(\log n)$ provide no advantages in any quantum algorithms. This rigorously rules out the possibility of implementing well-known quantum algorithms, including Shor's, Grover's, Harrow-Hassidim-Lloyd, and linear-depth variational algorithms. Then, we study the maximal entanglement that noisy quantum devices can produce under one- and two-dimensional qubit connections. In particular, for a one-dimensional qubit chain, we show an upper bound of $O(\log n)$. This finding highlights the restraints for quantum simulation and scalability regarding entanglement growth. Additionally, our result sheds light on the classical simulatability in practical cases.
[1]
Jingbo B. Wang,et al.
Complexity analysis of weakly noisy quantum states via quantum machine learning
,
2023,
2303.17813.
[2]
X. Waintal,et al.
Grover's Algorithm Offers No Quantum Advantage
,
2023,
2303.11317.
[3]
Isaac H. Kim,et al.
Experimental demonstration of the advantage of adaptive quantum circuits
,
2023,
2302.03029.
[4]
Daniel Stilck Francca,et al.
The modified logarithmic Sobolev inequality for quantum spin systems: classical and commuting nearest neighbour interactions
,
2020,
2009.11817.
[5]
Rodney Van Meter,et al.
ARCHITECTURE-DEPENDENT EXECUTION TIME OF SHOR'S ALGORITHM
,
2005,
quant-ph/0507023.
[6]
William H. Press,et al.
Journal of the
,
2023
.
[7]
P. Malliavin.
Infinite dimensional analysis
,
1993
.