I came across quantum computing the first time a couple of years back, I read about quantum cryptography even earlier. Resources on quantum computing looked suspiciously similar to my general Quantum Mechanics lectures at the university. quantum computing seemed too far away to me, to qualify as a productive use of my time and energy.
I completely lost sight of the topic, until spring 2022, when I found that you could rent an actual quantum computer via AWS. This got me wondering, what is the state of quantum computing, what are people hoping to gain from it, etc.
Why bother?
Computational problems in classical computing can be coarsely split into the complexity classes P (solvable in polynomial time, tractable) and NP (not solvable in polynomial time, intractable).
Many problems we care about are in the NP-class, we can only hope for “Okay” solutions, correct or optimal solutions are out of reach to us. These problems include among many others: the design of catalysts, protein folding or portfolio optimization.
Quantum computing promises better performance on many computational problems, and some problems that are in NP are not in QNP (quantum non-polynomial). The most famous example of an existing quantum algorithm is the factorization of large numbers, that could be used to crack SSH-encrypted communication.
High hopes and big players
I found a rapidly growing field, with research, big vendors, first hints of industrial applications and few community structures. Also, I found way too many shallow YouTube videos and newspaper articles, praising the new technology as solution to all known problems of mankind – I found it challenging to extract substantial knowledge from those.
What finally caught my attention, was the IBM Quantum State of the Union on YouTube. A big player with a track record for technology innovation, outlines an ambitious, yet (from the outside) plausible, road-map for their quantum developments. Their message: We’re on the brink of building application ready Quantum computers, and we intend to surpass the quantum equivalent of Moore’s Law (that took standard computers from a curiosity to a corner stone of modern civilization in just a few years) over the next couple of years. Alright, I want to learn more!
The bare basics
A very brief overview of what we’re talking about.
What are Quantums anyways?
Quantums are tiny physical objects, that cannot be accurately described by classical mechanics, but with quantum mechanics. In classical mechanics, quantities like energy, momentum, etc. can obtain any value. However, if you go down to objects like small molecules, atoms, electrons or photons, this theory doesn’t match observation. In quantum mechanics energy, momentum, etc. can only obtain certain values, the theory works remarkably well – so well, that we can use it to construct computing machines based on it’s principles.
What is a qubit?
A qubit is a quantum object, that holds information, and can be manipulated to perform calculations. Different research-groups and vendors utilize different quantum objects, like photons, ions, topological majorana particles, …
In quantum computing, quantum effects like superposition and entanglement are used to perform calculations in a completely novel way.
Superposition and Entanglement
Superposition and entanglement are two very useful quantum effects, that allow for completely novel logic operations.
- Superposition describes, that quantum objects can obtain in-between states that collapse to pure states on measurement. One common strategy to find solutions in quantum computing, is to use superposition by initializing qubits to all possible outcomes and use interference effects to amplify the correct result. On measurement, the amplified result will be obtained with a higher probability.
- Entanglement describes, that quantum objects can obtain a shared state, i.e. the objects are no longer independent of each other. Measuring one qubit will determine the state of the other qubit.
How are instructions performed on qubits?
This looks very different in the different implementations. In the case of photons, the instructions may be mirrors, polarizers and crystals; in the case of Josephson junctions, the instructions may be microwave pulses.
What is the significance of measurements?
Qubits can obtain states in superposition and in entanglement, however, we cannot read out the information on that level. On measuring our qubits, we end up with simple binary information 0 or 1. Once we measure a qubit, the inaccessible information is lost.
The goal of quantum algorithms is to increase the probability of measuring the desired result. Quantum algorithms are executed multiple times, to retrieve a probability distribution of the possible results.
Promises and Challenges
At this point, the current state of development is difficult to assess. There is certainly a sensation of a gold rush, part of it is hype, part of it looks very promising to me. The technology seems to be improving day-by-day, still many hurdles need to be overcome.
I believe, now the success stories are discretely written, we will be reading about in a few years time.
My Way forward
I personally want to learn more about quantum computing and get more hands-on experience in the field. I read one bad and two good books on the topic, currently, I’m working on the standard volume “Quantum Computation and Quantum Information” by Nielsen and Chuang. I reached out to several people in my home-town, and I want to attend a lecture this fall at my local university. I’m curious where this learning journey takes me.