Requirements for physical implementations
has classified the main ingredients that a physical system should possess in order to be a candidate for a quantum computer. They are:
- Quantum register: A scalable physical system with well characterized representing qubits that in turn compose the quantum register.
- Initialization: The ability to prepare the state of the register in a initial state.
- Universal set of gates: The ability to implement a universal set of logic gates.
- Low error and decoherence rate: High fidelity of gate operation, with probability per gate < 10−3 and qubit decoherence times that are longer than the gate operation time.
- Read-out: The ability to reliably measure the state of individual qubits the computational basis
With the knowledge of the basic requirements, here is an actual scheme for a physical implementation of quantum computation.
Nuclear magnetic resonance
Nuclear magnetic resonance
is a spectroscopic
tool that is used to get the structural information of molecule
due to the chemical shift
and Zeeman effect
on the resonant frequencies
of the nuclei
With a spin of S=1/2, the distance between the two energy levels is proportional to the magnetic field
strength. With the help of a coil
, which generates a radio frequency
(RF) magnetic field, it is possible to detect transitions between the different spin states.
The qubits can be realized by using the spins of freely floating molecules. A quantum register now needs several distinguishable qubits. We can realize that by the fact that different molecules have different Larmor frequencies. So spins whose Larmor frequency differs from the frequency of the radio frequency pulse are not affected by the pulse.
Pseudo effective pure states
Because detecting individual spins is extremely difficult in most cases, signals can
be detected only from an ensemble of spins, containing in the order of spins.
For the description of the mixed states a density operator is used. The corresponding average density operator corresponds to the sum of the unit operator (the totally mixed state) and a pseudo pure state. One big disadvantage of this
process is that one loses signal by destroying polarization. This loss of polarization, which increases exponentially with the number of spins in a register, restricts the usefulness of NMR quantum computing.
NMR signals are obtained in the time domain as a response of the system to an RF pulse. The system is virtually unaffected by the measurement. Of course, this is so because the system consists of an ensemble of many spins and not just a single particle. The observation of the spins is achieved through the Faraday effect. The polarized spin ensemble is a macroscopic magnetization. The generated signal is known as free induction decay (FID). This signal is analyzed in the frequency-domain after an Fourier transform. The frequency-domain contains the same information as the time-domain, but it makes it possible to distinguish between different transitions. Two distinct transitions usually have different Larmor frequencies. Their corresponding resonance lines are therefore separated in the frequency space.
One qubit gates are easily implemented using resonant RF pulses, which excite the corresponding spins. By using pulses with narrow excitation bandwidths it is possible to select single spins, and this perform operations on individual qubits.
The implementation of a NOT, for example, can be applied by a 180 degree pulse, which inverts the two states and . Two qubit gates can be implemented by combining one qubit gates with the spin-spin coupling. That way, for example, a Cnot gate can be implemented.
Because all spins have a different Larmor frequency to allow addressability for logical operations, their precision frequencies during detection also will be different. A Fourier transformation of the FID from such a system, therefore separates the contributions from different qubits in frequency space.
Practical implementations of the many-qubit systems are still in their infancy. NMR based quantum computers, which are
rather successful in terms of the number of qubits realized so far because they are not scalable. It is still very difficult to resolve the NMR frequencies of individual qubits and the measurement signal, as the number of qubits becomes larger than
a dozen. A more complete list of Quantum Computer implementations can be found at Quantum Computing
- DiVincenzo, David P. (2000). "The Physical Implementation of Quantum Computation". Experimental Proposals for Quantum Computation. .
- C. Adami, N.j. Cerf. (1998). "Quantum computation with linear optics". .
- Joachim Stolze,; Dieter Suter, (2004). Quantum Computing. Wiley-VCH.
- Ian Mitchell, Computing Power into the 21st Century: Moore's Law and Beyond. (1998). .
- Rolf Landauer, Irreversibility and heat generation in the computing process. (1961). .
- Moore, Gordon E. (1965). Cramming more components onto integrated circuits.
- R.w. Keyes, (1988). Miniaturization of electronics and its limits.
- M. A. Nielsen,; E. Knill, ; R. Laamme, Complete Quantum Teleportation By Nuclear Magnetic Resonance. .
- D. P. Divincenzo, (2000). The physical implementation of quantum computation.
- Lieven M.k. Vandersypen,; Constantino S. Yannoni, ; Isaac L. Chuang, (2000). Liquid state NMR Quantum Computing.
- Imai Hiroshi,; Hayashi Masahito, (2006). Quantum Computation and Information.
- Andre Berthiaume, Quantum Computation. (1997). .
- David R.~simon, On the Power of Quantum Computation. Institute of Electrical and Electronic Engineers Computer Society Press. (1994). .
- Seminar Post Quantum Cryptology. chair for communication security at the Ruhr-University Bochum. .