This question will be useful to many.
Can we collate a list of challenges in enabling Practical-Scale Quantum Computing?
Nathan Aw
Asked
Active
Viewed 138 times
-2
-
1) Too broad. 2) We're programmers, not physicists. – user2357112 Mar 02 '18 at 02:15
2 Answers
1
The challenges are dependent on the physical architecture. Pick a physical implementation and look for which of the DiVincenzo's criterias are missing. These would be the chosen architecture's challenges.
Exeko
- 56
- 7
-
For precision: Superconducting qubits and circuits based qubits (i.e. transmons) have issues with leakage towards states that aren't part of the encoding Hilbert space, making error correction much more complicated. . . Spin qubits mostly have issues with scalability, both in quantum dots and donor architectures (lack of physical space to implement all of the control and readout components). . . Almost all physical platforms still have issues with their 2-qubit gates fidelities. The "best" architecture so far is supercond. qubits; both IBM and Rigetti have few-qubits chips. – Exeko Mar 17 '18 at 17:49
0
First of all, I should mention that when people say they want a practical-scale quantum computer, what they probably mean is that they want a device that could perform a computation and yield a good result with a probability arbitrarily close to 100%. This can be achieved with imperfect qubits by using Quantum-Error-Correction algorithms, but it still requires somewhat decent quality qubits (and "decent" by today's standards is of the highest quality people do). There's a qubit quality threshold above which using those error correction algorithms becomes useful and reduces the logical error rate. Below that threshold, it is best to hope for no errors and not perform any error correction.
To answer the question: ""Are there physical implementations that fulfill the 5 DiVincenzo's criteria?""
Some architectures partially fulfill all of them, but not to the level needed to perform error correction & therefore enable practical-scale and reliable quantum computing. For example, IBM and Rigetti-computing (maybe google too? to be verified) have small-scale quantum computers using superconducting/circuit qubits, but each individual qubit is not that great (by error correction standards) and several of the qubits features are below Quantum-Error-Correction threshold. This means that performing a computation could yield an incorrect result. The probability to get a wrong result depends on a bunch of factors such as the coherence time of each of the individual qubits, single and 2-qubit gates fidelities, readout fidelity, etc.
Such a small-scale quantum computer does perform, in a sense, quantum computing. It is just not yet at the required reliability to do interesting applications in quantum chemistry or condensed matter (which are expected to be the killer-apps of a quantum computer).
Exeko
- 56
- 7