One might state the main goal of theoretical computer science as "study the power
and limitations of the strongest-possible computational devices that Nature allows us."
Since our current understanding of Nature is quantum mechanical, theoretical computer science should arguably be studying the power of quantum computers, not classical ones.
--- Ronald de Wolf
- single-qubit system
- $|0\rangle :=$ a state representing bit $0$
- multi-qubit system
- $|00\rangle := |0\rangle \otimes |0\rangle$
- $|00\rangle, |01\rangle, |10\rangle, |11\rangle$
- n qubit system has $2^n$ bases
Note:
- X is not, Y, Z is upto change of basis
- H is like sqrt ok identity, to uniform
- (Mention Universality) There exist universal quantum gate set that permits arbitrarily small error.
- $f:\{0,1\}^n \to \{0,1\}$ is either balanced or constant
- It requires $O(N=2^n)$ query of $f$ for a classical circuit to judge
- quantum: $O(1)$ query
Note:
- can briefly introducethe concept of complexity here (asymptotic time as function of n)
Gover's search is a quantum algorithm for the "unstructured search" problem.
The core idea is to see the input space as a combination of "Good" and "Bad" subspaces and do reflections (Grover's iteration) to make the probability of observing the "Good" part higher. To achieve a probability close to 1, about \(O(\sqrt{n})\) iterations are required.