Comments about the article in Nature: Quantum Computer Quest

Following is a discussion about this article in Nature News Vol 516 4 December 2014, by Elizabeth Gibney
To read the article select:!/menu/main/topColumns/topLeftColumn/pdf/516024a.pdf
See also: Poulin
In the last paragraph I explain my own opinion.


The article starts with the sentence:
In September, Google recruited John Martinis etc and set him to work on a notorious difficult task of building quantum computers: devices that exploit the quirks of the quantum world to carry out calculations that ordinary computers could not finish in the life time of the Universe
Next we read:
Even today after three decades of effort the best quantum computers in the world are barely able to do school-level problems such as finding the prime factors of the number 21.
If that is all then they have a long way to go.

Seveties Child

But in the quantum realm etc, then the subatomic laws that govern those particles make it possible for a given quantum bit to be both 1 and 0 at the same time.
The behaviour of physical processes is not governed by any law. For example, the evolution of the universe is not governed by any law at large scale nor at any small scale.
The fact if a quantum bit is both 1 and 0 at the same time should be established by obsevation or by experiment.
Next we read:
By extension the set of qubits comprising the memory of a quantum computer could exist in every possible combination of 1s and 0s at once.
Considering a register of 3 qubits.
Where a classical computer has to try each combination in turn, a quantum computer could process all those combinations simultaneous - in effect carrying out calculations on every possible set of input data in parallel.
That could be true. But how do you know that given two registers of each 3 qubits which are multiplied which each after which is equivalent which a register of 6 qubits, that the two values which when multiplied produce the number 21 are the numbers 3 and 7?
Shor's algorithm meant that in principle quantum computers could crack that encryption.
This should read as: What Shor's algorithm meant is that quantum computers can factorize large numbers faster than classical computers. That in turn means that you have to use even more larger numbers to have the same secure encryption scheme.
But achieving the ultimate goal - a general purpose, digital quantum computer that can be programmed to carry out any algorithm - has proved much tougher.
This sentence is very misleading.
  1. First of all you cannot program a quantum computer.
    You cannot write a program in a higher level language on a quantum computer which generates code which executes on the same quantum computer like you do on a PC.
  2. The problem is a quantum computer is a hardware device where all the hardware is involved in solving your problem. The larger your problem the more hardware you need.
    It is also important to understand that every problem requires more or less its own computer. As such you have to build your own hardware. In that sense a quantum computer is like a analog computer which consits of integrators, multipliers and adders which requires wires to configure for each individual problem.
  3. The best you can do is write a program on a digital computer which generates a list of instructions about what to do on a quantum computer.
  4. The largest number you can factorize on a digital computer is 1230186684560413680454744877 using 4 processors in parallel in roughly 1.2 day. See: Quantum Factoring Performance Evaluation - Using Parallel Processing - VB2010
    IMO this problem is an almost impossible task to solve on a Quantum Computer and by the time you can PC's will again has increased in performance.
Next we read:
Thanks to a quantum phenomenon known as the Josephson effect, electric currents flowing around tiny loops in such circuits can circle both clockwise and counterclockwise at once, so are perfect for representing a qubit.
They are only perfect if the chance of measuring a 0 or a 1 is identical.
If you have two qubits if the chance of meauring 00, 01, 10 and 11 is identical etc.
However that is not the only problem. The most difficult part is to perform logic operations on the individual states of the qubits i.e. registers.
On a digital computer "and" and "or" operations between individual bits are easy to build and to test, (because only electrical signals and clock pulses are involved). On a quantum computer the same is very difficult on the qubits based on the Jospson effect.
Such circuits are tricky to implement, says Martinis. "You have to work many years to figure out all the physics"
I can only agree.

Onwards and upwards

Their main tasks will be to figure out how to fabricate large scale qubit arrays, how to control the quantum computation and to read out the results and how to connect up the quantum circuitery to classical electronics that reside on the same chip.
A diferent problem is how to initialize the state of a ship using classical electronics i.e. a PC.
You have the same problems of what is called a hybrid computer which is a combination of a digital and an analog computer.
Each extra qubit increases the complexity of the hardware.
Correct. When you make one register larger all the hardware involved has to be updated.
Matthias Troyer estimates that working out how to wire up, manipulate and fabricate qubits in bulk will be a US$10-billion problem. That poses a crucial question he says. "Why should one do it?"
Correctly. What are the benefits beside speed and what are the disadvantages.
See also Reflection 1
The two classical examples, code-cracking and searching databases are not good enough says Troyer
  • Shor's algorithm will require thousands of qubits to do any serious factorization.
  • Quantum computers may search databases faster, they are still limited by the time it takes to feed the data into the circuit which would not change
I expect that searching large data bases require millions of qubits.
Troyer thinks that a much more fruitful application for the near future is the modelling of electrons in materials and molecules.
For example: the iron sulphide in the ferredoxin proteins that are involved in the nitrogen fixation in plants.
To do that you need two algorithms: One to perform such a task on a PC and a second one on a Quantum Computer. Both should give the same answer.
The second one is the most difficult.
Troyer thinks that a quantum computer could help to design a catalyst that would be much more energy-efficient than the current ones.
You can always think about it but it should also be realistic.
At the end of the article we read:
"Google created a new name for scientists working on the hardware effort 'quantum enigineers'," says Martinis. "This is a dream "

Reflection part 1 - Quantum algorithms

In order to solve any problem on any PC you need an algorithm, a document that the describes the computations and the decisions step by step sequential in a structured mathematical language.
Two things are important. For a Quantum Computer the whole situation is much more complex. The whole idea of a quantum computer is to perform the calculations as much as possible in parallel. By preference as little as possible is done in parallel. One of the major questions to investigate is if this is possible for all problems and if there are advantages. The result will most probably be that only a small range of problems there will be benefits.

To use a Quantum Computer as a tool to generate random numbers is extremely easy. The whole problem boils down to what extend you can measure the state of a qubit with a 50-50 chance to be equal to 1.
At the same time it is very difficult to devellop a quantum algorithm to calculate pi and to implement such an algorithm. At the same time "Why should one do it?" when the number of digits behind the comma is not high.

To perform parallel programming on a classical computer is rather difficult. To do a search using parallel programming is easy because you can divide a data base in for example 8 different independent subsets. To rearrange a data base you can use the same technique to improve speed. The final part of combining the subsets is also easy when the whole database is what is called core-resident. For physical systems like galaxy simulations this is very difficult because everything depents on each other.
In all these case to devellop quantum algorithms is difficult.



Created: 23 July 2014

Back to my home page Index
Back to Nature comments Nature Index