Google – Quantum leap?

Quantum computing earns a place on the far horizon.

• Digging beneath Google’s characteristically effusive claims reveals a solid advance that significantly increases quantum computing’s claim to be the successor of the silicon-based systems in use today.
• Google is laying claim to “quantum supremacy” which is considered to be the watershed moment when a quantum computer can outperform the most advanced silicon-based system available today.
• At a very high level, the evidence is convincing.
• Google’s quantum computer was able to perform a calculation in 20 minutes that would have taken the most advanced classical computer today 10,000 years.
• As ever with these sorts of things, the devil is in the details and there are two big shortcomings with this research.
  • First, simulation: the veracity of this statement is untestable because Google was unable to verify that it would indeed take the classical computer 10,000 years to perform this calculation.
  • This is for obvious reasons as otherwise, the experiment would be running until the year 12019.
  • Hence, the calculation was scaled down to a reasonable time frame for the classical computer to gauge the difference between the two systems.
  • The quantum computer was then given the original calculation and an estimate was made based on the scaled-down calculation, how long the classical computer would have taken to solve it.
  • While I don’t really doubt the veracity of this claim, it is at the end of the day an estimate and not hard fact.
  • Second, calculation: the task that was tested was specifically designed for the quantum computer.
  • This task was random number generation and then ordering them into a specific pattern designated by the algorithm.
  • This means that any other practical tasks such as running highly power-intensive deep learning training systems will not run on this system.
  • Consequently, as it stands today, for practical purposes the quantum computer is useless.
• For these two reasons, I would question the bold statement by Google that quantum supremacy has actually been achieved.
• The quantum computer is incapable of performing the practical and useful tasks performed by today’s supercomputers and therefore fails the requirements of quantum supremacy.
• When a quantum computer can train massive neural nets in a fraction of the time and much more cheaply in terms of power consumption then quantum supremacy will really have been achieved.
• Regardless of this fact, what Google has achieved is significant as the improvement in speed is so substantial that it clearly demonstrates the potential of quantum computing.
• This will galvanise research efforts and, in all likelihood, trigger an increase in funding.
• Google has firmly placed quantum computing on the horizon, but the journey has many many years yet to travel.
• Silicon is here to stay for the foreseeable future.