Proponents of quantum technology believe its will change the world. Others remain skeptical, as they do of technologies like fusion energy.
Speaking at a quantum developers’ forum, IBM Distinguished Engineer Jan-Rainer Lahmann retraced the history of quantum computing, reviewing IBM’s hardware and development roadmaps and describing the ingredients of “Raspberry Pi quantum”.
The history of quantum computing goes back four decades to a conference where the Nobel laureate Richard Feynman introduced the idea of simulating quantum mechanical systems on a traditional computer. At the time, this required a significant computational resources. Even with Moore’s Law scaling, it was clear to Feynman and many others that the road to quantum computing needed to be pursued. “What if we built completely different kinds of computers that made quantum mechanics’ effects such as superposition, interference, entanglement, directly accessible and controllable?” Lahmann recalled Feynman as asking.
Lahmann continued: “With such a different kind of computer, it should be much easier to simulate quantum mechanical systems. I think this idea is very clear, and it makes perfect sense.”
Since then, many scientists and engineers have pursued various approaches to building actual quantum computers. Feynman’s basic idea was that a quantum mechanical system, with several subsystems, for each qubit, provides as many traditional bits as would be needed on a traditional computer to express that state of a quantum mechanical system. For example, 2 qubits are equivalent to 512 bits, 10 qubits are equivalent to 16 kB and so on with exponential growth. Also understood at the time was how difficult it was to build large computers that could handle qubit demands.
“If you have a quantum mechanical system, you need a huge traditional computer to simulate the same things; if you have a traditional computer, then you can express this amount of information on a quantum computer under certain conditions,” said Lahmann.
Increasing the speed of a quantum computer only makes sense for very specific problems. In an example, Lahmann described how long a quantum computer and a traditional computer would take to multiply two numbers. P and Q are integers with 2,048 bits. On a traditional computer, it takes a few milliseconds. And on a fairly small and noisy quantum computer, it would take an estimated 75 seconds.
But as Lahmann noted, a similar but much more complicated problem illustrates the potential and speed of quantum computers. “We don’t want to multiply two numbers, we want to factor a large number. So we have a number of 2,048 bits and we want to derive the prime factors of that number. This is the core of our two big asymmetric encryption schemes. This takes a long time on the traditional computer, on the order of years – this takes a couple of billion CPU cores on a traditional computer.”
Citing Peter Shor’s quantum algorithm, if “we have a large enough quantum computer, this could be reduced to a few hours. That vividly