Quantum computing – Under starter’s orders pt. II

D-Wave makes a real(ish) world demonstration.

• D-Wave has demonstrated that its quantum computer can solve a “real-world” problem and in the process found that it completed the task somewhere between 1m to 3m times faster than a regular supercomputer powered with silicon CPUs would be able to do (see here).
• This kind of speed is not a new revelation but the major drawback to date has been that these machines have not been able to do anything that would remotely approach a real-world problem.
• The “real-world” problem that the researchers decided to take on was “exotic magnetism” which occurs in quantum magnetic systems.
• In 2016 Thouless and Haldane won the Nobel Prize for physics by using advanced mathematical techniques to explain the exotic magnetism and phenomena of superconductors, superfluids and thin magnetic films.
• In practice, these techniques helped explain why certain materials behave the way that they do when they are cooled to almost 0°K (-273°C), the temperature at which molecules stop vibrating and most of the established physical and chemical rules go out of the window.
• The D-Wave supercomputer was able to reproduce the calculations of these techniques and do so over a million times faster than a regular supercomputer would be able to do it.
• Effectively, D-wave built a mathematical model of the chemical structure and had the compute power to simulate how it would behave under certain circumstances.
• This is being held out as an example of how quantum computing can have real-world uses but I think that this is stretching the realm of credibility somewhat.
• This is because when people hear “real world” they think algorithms that can match videos to viewers better than Tik Tok or drive vehicles rather than a quite obscure branch of quantum physics.
• Furthermore, the system had to be specifically designed to solve this task meaning that we are still pretty far away from a simple system that pretty much anyone can program to do anything.
• However, it does represent a step forward and D-Wave is a going concern in that it sells quantum systems largely for research purposes and is beginning to allow 3^rd parties write programs for its quantum systems in the cloud.
• There are still many different techniques being used to create quantum computers and we are still very far away from a standard or anything that could be close to replacing today’s silicon-based systems.
• Hence, silicon still has life left in it and the horizon for when it will no longer be able to reduce costs, power consumption while increasing performance is still quite far away although it is rapidly slowing.
• This combined with the plethora of experimental methods of quantum computing and the Chinese favoured optical computing systems means that there is no clear replacement for silicon on the horizon yet.
• This won’t happen until one system is developed that is demonstrably better than all of the rest and some standardisation begins.
• There are also a host of other technical problems to solve such as the requirement to cool the processors down to almost absolute zero which ensures that these machines are expensive, bulky and consume huge amounts of power.
• This is why I continue to believe that silicon still has many years to run and why I still keep quantum computing on the same time horizon for real commerciality as nuclear fusion.
• To be fair, quantum computing is making more progress on this front, but silicon is going to remain the main way forward for the foreseeable future.