Quantum computing: Hype or reality? OVH says businesses would be better off prepared
Understand the technology before you are taken by surprise, argues CTO
Euro cloud biz OVH recently announced the purchase of its first quantum-powered system as part of long-term plans to build a quantum environment for developers and other users.
The company has a pragmatic approach, viewing it as a technology that has a long way to go before it will live up to the hype, but nevertheless important enough that organizations should do their homework and be prepared for when quantum is more mature.
The Register spoke recently with OVH chief technology officer Thierry Souche, who discussed his views on quantum computing and the company's roadmap.
Souche disclosed that he took some time out to get trained up on quantum and how it works, and said it is something he strongly suggests that other executives should consider doing.
"I was always struggling to understand how a quantum computer would work, and why it was such a step change versus classical computing," he said.
"So taking at least a day and a half or two of training on just doing the basics, the quantum Hello World if you like, helped me gain a lot of insight on why it could be so different, but also the limitations associated with the technology, and especially why it will not be a do-everything type of technology and you would still need classical computing."
His take on quantum tech, following the training, is to think of it almost as the reverse of how a conventional computer works, right down at the lowest level of the hardware. And that's where quantum computers basically are at the moment: still at a primitive level, comparable to where classical computers were more than 50 years ago.
"When you consider a classical computer, you have a piece of silicon with the various gates which combined will define a circuit implementing some logic. So the static thing there is the gates, millions of them, and the flow is of electrons," Souche said.
"Quantum computing is actually the reverse. If you consider one of the technologies, cold atoms, what is static is the information you store. And the dynamic part is how you apply a gate on the atoms. Applying a gate means actually getting one specific atom into a specific level of excitation."
In this example, that means applying a beam of energy to that single atom, and the way it is applied – the level of energy, spectrum, duration – serves as the gate.
You could very easily make a decision to invest in a dead end.
"So you dynamically flow in logic, while what you keep static is the memory, which is completely the reverse of what you do with a classical computer," Souche said.
Whether or not you find this way of looking at quantum helpful, Souche believes that the kind of very basic training he sat through would assist decision makers in all organizations to at least grasp how different quantum is, why they need to think about it differently, and to try to understand how it may be put to use in their business.
"While the technology will be great and nice, it's still in the experimental phase, and it will not cover each and every problem, especially if the nature of the problem you want to solve is not amenable to a particular quantum technology. And if you don't grasp that, you could very easily make a decision to invest in a dead end," Souche told us.
These quantum technologies use different ways of representing qubits, or quantum bits, with a variety of them under development, including trapped ions, quantum dots and superconducting circuits. Some of these technologies may prove to be better suited for certain applications than others.
"I think a lot of decision makers and their teams need to go through that journey in order for them to assess whether there is a competitive advantage to go for quantum technology, or whether classical computing will do the job just as well. They need to know where they should never ever try to use quantum because it will deliver nothing," Souche said.
"What is sure is that there are very few people trained. So if you wake up too late, you will simply find no one available to help you assess whether this is or is not the technology of choice for you. But if you start training your people now, you have a good chance to have staff in-house that can do the assessment for you. We already know that for some industries it is likely to be a game changer."
- India gives itself a mission to develop a 1000-qubit quantum computer in just eight years
- Amazon: Diamonds are a quantum network's best friend
- D-Wave hello to another quantum pioneer warned over possible delisting
- Microsoft wants you to build quantum apps in Azure, the cloud that's both up and down
Areas where OVH believes there is a deep fit for the technology include chemistry and modeling molecular folding. "When you need to model actual real life molecules with thousands of atoms or more, you can't do this with conventional computing. But we think that in a few years' time, you will get that with a quantum computer," Souche told us.
However, "when you think about your day-to-day business applications like accounting or CRM, don't dream about quantum, I don't think it will help you," he added.
OVH still believes that organizations need to start at least looking into quantum, and how it may affect their line of business in the future. "Don't be taken by surprise" is the key takeaway.
This is why the company is itself investing in quantum, Souche claims, in order to help customers with this process.
Don't wait especially on training your people. Because they will be the biggest asset in the transformation to come.
"Today, we already offer quantum-as-a-service, but it is emulation. And we offer that for a range of technologies. The emulation, which is backed by powerful conventional computing, already provides the ability for a lot of companies to assess the type of algorithms that they might want to test," he said.
"We're working with several companies, and to begin with we will expose their machines in their datacenters through our cloud. And then we have plans, but I cannot yet say more, for several types of quantum machines are installed in our datacenters. But that's more for the future because they're not ready," Souche added, explaining that some require cryogenic cooling "and outside a lab environment that is not a good thing."
Looking ahead, the CTO thinks that at least five or six quantum technologies will emerge in the near future, and this may be whittled down to maybe one or two eventually.
"But no one can currently give me a plausible explanation of why any might be better off later on than any other. They all come with trade-offs, and none of them are ready to scale tomorrow. There are plenty of trade-offs always. So it's still very speculative," he said.
"But I would still say don't wait. Don't wait especially on training your people. Because they will be the biggest asset in the transformation to come."
Gartner VP analyst Chirag Dekate told us that there is so much hype around quantum that many in the industry end up polarized between quantum skepticism and those that push it as some sort of nirvana that will solve all problems.
"With uncertain macroeconomic conditions, this polarization is likely going to get worse, as companies chase after shrinking pools of money in their quest for survival," he said, stating that the majority of quantum startups will struggle to make it through the market cycle that has already started.
Dekate said that Gartner is aware of more than 300 enterprises that are in different phases of their quantum journeys, and while most are just getting started, a growing minority are already investing millions in creating IP and building relationships with relevant vendors and service providers. ®