Mar 14 2023

Update on Quantum Computers

There has been a lot of quantum computer news since I last wrote about the topic. But this is still a technology that is slowly advancing in the background, while actual applications have been limited. There is a threshold effect at play – at some point, quantum computers will be powerful and practical enough that there will likely be a rapid adoption at the scientific, governmental, and industrial level (these are not for personal use, at least not anytime soon).

Quantum computers, as the name implies, exploit properties of quantum mechanics, rather than classical physics as do traditional computers. Digital computers, like the one you are using now to read this blog, are largely based on the bit – a bit is the smallest piece of information, either a 1 or a 0, in the binary language of computers. Computer technologies require some physical property that is very small and that can quickly and accurately be changed between two states, with these two states representing a 1 or 0. This can be done in various ways, but mostly exploiting electromagnetism.

Quantum computers do not use bits, they use qubits. A qubit exploits a quantum property (superposition) that can be in any state between 1 and 0, inclusive. Each qubit, therefore, has the potential to store orders of magnitude more information than one bit. Qubits are then entangles with each other and this entanglement produces constructive and destructive interference which can magnify the probable correct answer (gross oversimplification, but follow the links in my previous article for more details if you are interested). A working quantum computer requires a physical property at the quantum level that can be in superposition and entangled, and connected to every other qubit. There are still major hurdles to overcome, one being error correction.

While the size of quantum computers are slowly and steadily growing, scientists are still researching new basic methods for building them. IBM is set to release the first >1,000 qubit computer, Condor, this year. They also plan to build a platform that networks together multiple Condor processors.

Meanwhile new methods of making qubits are frequently being reported. The latest is a study that uses two lasers to alter the spin of nuclei. One of the researchers, Paolo Cappellaro, says:

 “This novel coupling mechanism enables their control and measurement, which now makes using nuclear spins as qubits a much more promising endeavor.”

That’s the key – the technology needs to both control and measure the property that will store the qubit information. At this point there are various options and it’s too early to tell which approach will prove most practical. But we need more than qubits, we also need integrated circuits. There has been progress on that front as well. In 2022 an Australian company announced they produce the first quantum integrated circuit. They did this in silicon, which is helpful because we already have an industrial base with silicon integrated circuits. They anticipate it will be another five years before there are commercial products using their quantum circuits.

Why should we care? While quantum computers will not replace your desktop or smartphone, they are useful for certain kinds of intense processing. To put their power into perspective, a quantum computer with 1 million qubits (still a ways off) could do a calculation that would take classical computers trillions of years to complete, even if they were made from a billion galaxies worth of matter, and it could complete that calculation in minutes. This basically means they could perform calculations that are impossible for ordinary computers.

Two of the most anticipated applications for quantum computers are encryption and modeling. A quantum computer could break any existing encryption. Only another quantum computer could defend against decryption by a quantum computer. In the world of cybersecurity, whoever gets a working quantum computer first wins. This is the equivalent of nuclear weapons in conventional warfare. “Mr. President, we cannot allow a quantum computer gap!” (bonus points if you get that reference)

For science and industry modeling with also be important. For drug development, for example, quantum computers could to protein folding modeling that no standard computer could do. Material science, climate science, brain mapping, and other industries could be transformed by quantum computers.

Now, here is where things get interesting – imagine a complex AI running on a powerful quantum computer. I don’t think we can even really predict at this point what this will mean. Quantum computers will not automatically make AI smarter or sentient. There are still other variables, such as the algorithms themselves and the training data. But quantum computers create the potential for AI applications that are many orders of magnitude more powerful that what we have today. We will discover the implications and applications of this when we get there, but it’s interesting to think about.

With IBMs 1,121 qubit computer, and the introduction of quantum integrated circuits, it seems like we are about 5 years or so away from the transition of quantum computers from pure research to commercial applications. As always, this is hard to predict, but it does get easier the closer you get, when applications are “in the pipeline”. We also are still not sure where China is with their quantum computer research, and they may surprise us. But it seems likely, by the 2030s at least, that quantum computers will be a technology running in the background of our society and having profound effects on our technology and lives.


No responses yet