Everything you need to know about quantum (but were afraid to ask)

Hiding in plain sight, in the window of IBM’s Waterloo offices in London’s York Road, is a machine that could soon change the world.

Although it resembles something from an episode of Doctor Who, it is actually the guts of a decommissioned quantum computer.

As machines go, it’s an opulent affair – but most of its chandelier-like copper wire and stainless-steel form is an elaborate fridge – it’s the little encased chip at the bottom, which looks like a SIM card – the Quantum Processing Unit (QPU) – that looks set to solve some of the world’s most complex problems.

While AI and large language models appear to have set the world on fire, quantum is still something people think of as humming away in the fridge of some basement, manned by technicians in white coats.

IBM however, a firm considered the world leader in quantum computing, believes that it’s something that industries of all sizes need to start engaging with now.

We’ll get onto the why in part two of this report, but first let’s take a closer look at the technology behind these systems – which TI got to witness close up after viewing a workable model earlier this month.

The computer is encased in glass and housed within a darkened room at IBM, which I was ushered into by one of its distinguished engineers and quantum ambassador, Richard Hopkins.

The machine chirps, which is a surprise. Shut your eyes and you could be inside an aviary.

“I always think of it as a slightly asthmatic steam engine,” Hopkins counters, as I share my thoughts, he explains that this is the sound of the compressor.

This compressor, he adds, extracts heat from the machine. It takes about 24 hours for a quantum machine to cool down to a working temperature.

Quantum computers, while consuming relatively little power (the compressor uses as much as a domestic washing machine), need to be kept at absolute zero to operate.

As Hopkins explains: “Everything around the QPU is there to divorce the chip from the rest of reality, the area around where that chip is working is about 13 millikelvin – which, as far as we know, is the coldest place on earth. It’s 30 times colder than outer space.”

Qubits

Qubits are the basis for quantum computers. They transcend the 1 or 0 binary code that classical computers are based on because, unlike bits, qubits can exist in multiple states simultaneously. That is to say, they can harness quantum properties to be both 1 and 0 at the same time.

It’s mind-blowing to think about, and involves some high-level quantum physics, but what’s useful for businesses to grasp is that it’s this property that gives qubits the potential to process exponential amounts of information.

Experts believe that the advantage quantum possesses over classic linear computers, is that the former is better at producing accurate answers to problems containing a large number of interrelated variables. It doesn’t estimate them – it can determine the absolute answer.

With just 275 qubits, it’s possible to hold more computational states than there are atoms in the universe.

Consequently, Hopkins observes that early use cases for quantum computing have involved “anything that requires complex optimisations or complex decision making with more than 100 elements within them”.

The processing power of quantum increases as you add more qubits. “In a quantum computer,” notes Hopkins, “add one more qubit, entangle it and you’ve essentially doubled its capability.

To date, IBM’s biggest quantum chip on the market is Osprey – which possesses 433 qubits. Given the bird-like sounds these machines emit, it seems fitting that all the company’s chips are named after birds of prey.

Osprey contains more than triple the 127 qubits of the Eagle processor released in 2021. Next up is the 1,121 qubit Condor, which IBM says is on track for delivery later this year.

Super vs quantum

A good thing about quantum, according to Hopkins, is that most of the algorithms are scalable.

“So if you want to solve a problem you can do it with really small numbers of qubits and gates and then prove that it works with a small example, and build up,” he says.

“That’s why we already know that quite a lot of quantum algorithms work – but we are waiting for the machines to have enough qubits to execute coherently for a long enough period to prove that it can go beyond what a super computer can do,” he notes.

According to reports in various journals, we’re starting to see some isolated examples of quantum supremacy – that’s the moment when a quantum computer does something that goes beyond the reach of any practical classical algorithm.

Google claims to have done so in 2019. Various scientists in China claim to have proven this in 2021.

One issue with quantum computing, however, is that the system can become corrupted by errors – there’s still a lot of “noise” as they say, in quantum circles.

However, even on the lower qubit Eagle, IBM has proven quantum’s usefulness in calculating a real-life problem with a higher degree of accuracy than a supercomputer, as reported in science Journal Nature this June.

“That shows how we’ve now tackled a scientific problem and we’ve produced results beyond the reach of any bog standard super computer on the planet,” Hopkins enthuses.

“Now we’re looking for it to solve inarguable real world problems in a commercial setting,” he says.

The cloud

IBM has installed its quantum computers in Germany, Japan, and in the US at Cleveland Clinic, with more in progress in Korea, Canada and Spain.

For the average business however, it’s not possible to maintain one of these units – even in a managed data centre. However, in 2016 IBM reached a milestone when it made quantum computing available in the cloud.

Today, the computing giant currently operates a fleet of 24 quantum systems on the cloud, a move that has enabled outside researchers and developers to explore possibilities remotely.

Another step forward was the opening of IBM’s Quantum Computation Centre in 2019, for commercial and industry use, which now boasts over 450,000 registered users and 250 members.

Users span a range of industries, including healthcare, automotive, banking, and finance. HSBC and Vodafone are flagship partners.

They use an open-source development platform for quantum computing developed by IBM called Qiskit, which enables institutions to collaborate and share their approach to complex problems, ideas and research.

It’s interesting that when we talk about the world’s yet-to-be-solved problems, they don’t all involve astro physics, as you might imagine.

According to Hopkins, it could be issues in logistics: working out the fastest way to off load containers from a ship and onto the port; or, due to the worldwide shortage of liquid nitrogen gas at present, and the limited amount of tankers, working out the best ways of optimising its distribution.

And then there’s banking. “One of the long-term examples we use is optimising the cash position of banks,” says Hopkins.

“At the moment it’s done through approximation because they don’t really know how much cash banks hold – but if they did know precisely and if they knew all the time then the chances of anything going badly wrong are massively reduced.

“Measures like that could make a huge difference to the world economy. It’s one of those problems that’s a long way off, but that’s a really useful thing to be able to do.”

*Now read the second part of our report, which addresses the issues of quantum safety, skills and training, SME engagement and what happens when you fuse quantum with AI