Quantum Computing Competes With Every Other Enterprise IT Priority

Summary

Heather West, a research manager at IDC, focusing on quantum computing, is interviewed by Yuval Boger. Heather and Yuval talk about where quantum computing is prioritized relative to over enterprise IT opportunities, the potential integration of quantum into the HPC data centers, full-stack vs. best-of-breed vendor selection and much more.

Listen below, or get the full transcript at the Quantum Computing Report

Transcript

Yuval Boger: Hello, Heather, and thank you for joining me today.

Heather West: Well, thank you for having me. This is a great opportunity.

Yuval: So who are you and what do you do?

Heather: So, I am a research manager at IDC. I work in our Infrastructure Systems, Platforms, and Technologies group, and my primary research focus is quantum computing.

I lead our quantum computing research at IDC, but I'm very fortunate to work alongside our group vice president and general manager, Ashish Nadkarni, and research vice president, Peter Rutten. Between the three of us, we launched the quantum computing research at IDC in 2018.

And since then, we've been on this wonderful ride of learning about quantum and meeting and working with people within the quantum ecosystem.

Yuval: Guests that come on this podcast are typically from the quantum industry, so they think that quantum is the most important thing in the world. But you speak with enterprises that have other issues as well.

Where does quantum rank in their priorities?

Heather: So, yes. I also contribute to research on workload infrastructure, storage and server systems that support the various workloads, and I assist Peter on some AI research as well. So we get to talk to a variety of different end users, and even within our own quantum research, we speak to a variety of end users.

Had we spoken back in 2018 when we began researching quantum, there were very few enterprises and companies that were even thinking about what quantum computing is. It was still more of an academic technology. But over the course of the last four years, what we began to see is IT developers first taking an interest in quantum computing, because they saw it as a way to solve problems that they just couldn't solve using their current class of compute technology.

And that's where a lot of these quantum initiatives within enterprises, the early adopters, really started. They began to research the technology and figure out how they could start experimenting with it in their own IT department, and then began making the business case that filtered up to the executive level.

More recently, what we've begun to see is another transition, where executive-level stakeholders within enterprises or companies are starting to understand the value that quantum computing will bring in the long term. As a result, we're now seeing interest at the executive level, with the focus of saying, "Okay, this is something long-term, but it's a technology that we should start thinking about and experimenting with today so that we can gain the value once that fault-tolerant computer is realized later."

Yuval: What does the typical experimentation look like? Is that a two- or three-person team and a little bit of cloud time, or is that a more systematic effort to identify use cases, to contract with vendors, and so on?

Heather: That really is very dependent on the company. Some companies understand that this is a long-term endeavor, and they understand that experimenting with quantum, doing it right, really identifying those use cases, and starting to build those algorithms and applications is a process that should be well mapped out, so that they get to where they want to be by the time we hit that fault-tolerant machine.

As a result, we see those sorts of companies investing more. And when I say more, we're not talking 50 to 75% of their IT spend. We're talking small proportions, less than 10%. But they're investing more. They are establishing a quantum computing champion within the organization itself…

Or developing a quantum research group, which starts to educate both themselves and others within the organization, whether it be the IT department, who will need the skills and other resources and want to know how this changes things and how it can benefit them, or the executive level, who might need a better understanding of the business value and what the expectations are.

Some executives are still really focused on short-term ROI versus the long term. So in that respect, it's a very well-defined process for how quantum is being adopted, well, I didn't mean to say implemented, but experimented with. And then we have other companies, similar to what I mentioned in the beginning, where the IT developers really understand that this technology is going to be an industry disruptor, that it will help them solve those intractable problems that are just out of scope for classical compute at this particular point in time.

And as individuals, maybe one or two people within that IT department, they're taking it upon themselves to self-educate or experiment, and then trying to convince others within the company that this is something that really should be a focus, or an initiative started within the company.

So there's really a continuum of companies, from those who are really well structured to those with just a couple of individuals within the IT department trying to drive the process.

Yuval: You spoke about a quantum champion, and I'm curious if you see a pattern. Where does this quantum champion sit?

Is that in the IT department? Is that in the business unit? After all, quantum could potentially impact many areas. Or maybe it's in the security department, where they worry about quantum security. Where would that champion sit?

Heather: Well, the best example that I've heard when talking to various end users is a quantum research group that's essentially a group of champions representing all the different departments you spoke about: the business area, the executive level, security.

Because once stakeholders within the organization start to realize what quantum is, and maybe they don't understand it in technical terms, digging down into what a qubit is and talking about pulse control and all that, but once they start to realize what the potential advantages could be, then everyone has a problem they want to try to run on quantum.

And as we all know, quantum computing is complex. There's definitely a learning curve that goes with it. So by including members of these various stakeholder groups, it helps to really understand where some of the largest problems exist, and which problems would be best suited for quantum, either to demonstrate business value in the nearer term if possible, or which use cases are more in line with the mission and the future development of the company.

So that's how I've seen it used best: really creating that quantum research group to lead as a champion unit. Because, as you said, when you have one person trying to represent the whole company, it can become problematic trying to understand all these other areas and their priorities.

Yuval: I'm sure there are many companies that rely on IDC as a trusted advisor, and I'm wondering what questions you are fielding from customers that want answers from IDC. And I'm guessing some of these answers come in the form of general, widely available reports. What reports are coming from IDC about quantum?

Heather: So we're in the process of putting together what we call an IDC MarketScape. We're looking at quantum hardware vendors with ours, and it looks at the hardware vendors within the quantum market to determine what their role is in this particular market segment.

So it's not necessarily saying vendor A is number one, vendor B is number two, vendor C is number three, and so forth, and ranking them per se, but where do they stand? Is there a vendor that's a particular leader within the market? Are there some up-and-coming vendors that maybe haven't advanced to the point where they're developing a full stack, but the technology they're developing and introducing into the market is innovative?

This is a hardware vendor that should be kept an eye on as they start to progress and make more announcements. So that's one that's coming; we're in the process of developing it at this point in time. And then we will be releasing our second quantum computing forecast, I believe at the end of Q1, which will give an update to the forecast we provided in November of 2021 regarding market growth and where we expect it to be within the next five or so years.

So those are a few of the big reports that we expect to see coming out.

Yuval: Do you see companies engage directly with quantum hardware vendors, or do they prefer to go through a cloud provider like Amazon Braket or Azure and then be able to work with multiple providers?

Heather: Both. Some of the companies that we speak with are still a little bit wary, or I shouldn't say wary, but… well, they're all in the experimental process, but some of them are unsure which type of technology would be best suited for the problems they're going to solve, or for their industry.

Like I said earlier, the fault-tolerant quantum machine is years out from being realized. In that time period, we could see new technology come about that we haven't seen yet, and that could bring new advantages. It could be better suited for solving particular types of problems for particular industries.

We just don't know. So by investing in, subscribing to, or working with a particular quantum hardware vendor, some companies are concerned that they will get into vendor lock-in, or maybe a generation lock-in, because they're only working with one vendor and the hardware being developed by that vendor.

For those organizations, we hear they prefer to work with some of these cloud service providers, because it gives them exposure to multiple types of quantum hardware being developed. It gives them access to the different types of qubit technology.

So for instance, you might work with one cloud service provider that offers access to a superconducting, as well as maybe a neutral atom, as well as maybe a photonic system, and those technologies work really differently. But in addition to those technologies, they also gain access to all the tools and resources they need to develop applications, and perhaps some consulting services to help them identify use cases and walk them through the process of learning quantum or being quantum-aware.

And they also get access to hardware-agnostic software that allows them to transition very easily between one quantum hardware type and another. As a result, they don't have to risk investing in an algorithm or application that works with one particular system but won't work with another, or with one generation and not the next.

Yuval: Have you seen companies that are disappointed in quantum, that basically say it's too early, it's 30 years away? I mean, they used to say that about nuclear fusion, and look what happened, right? But do you see companies that are disappointed in quantum, or not yet?

Heather: I wouldn't say disappointed; I would say hesitant.

I think it's too early to say disappointed because, again, the technology is so early on that it's really just this exploratory stage of figuring out what quantum is. That's a big question: What is quantum? Why should I be investing in it? What use cases or problems are most applicable to it?

A lot of the companies that we speak to are still in those very early phases. So I don't think disappointment is necessarily the right word. I definitely think viewing it as more of a risk, and asking why they should be looking into it, is more of what we hear.

Because again, it's still really early on. The technology is still very unstable. There's a lot of work that needs to be done. There's a lack of consensus as to which technology will be the winner, if there is a winner, when quantum advantage will be realized, and so on. So in that respect, that's more what we're hearing.

Yuval: Do you see a big push to integrate quantum into the high-performance computing center?

And do you see that primarily as a cloud-based integration or on-premise?

Heather: We're hearing more and more about the potential to integrate it into HPC centers. I would say that conversation really started about six months ago, maybe a little bit longer than that, but not quite.

Before that, it was just, "Oh, what is quantum good for? What are these use cases?" But we have started to hear more and more about it, and there are a couple of different approaches. We are hearing from some organizations that are really concerned with security and performance, and as a result would like to have an on-premise system.

But obviously, as you and I both know, these systems are really, really early on. They're very complex to maintain and run, and that provides some challenges in itself. And a lot of the hardware vendors themselves, because it is still early on, would prefer to give access in the cloud until they are able to resolve some of the issues around the hardware itself.

But there are some vendors out there who do have on-prem options. Still, we are hearing from end users that they prefer using the cloud for many of the same reasons I just spoke about, and also because some of these cloud service providers are making it really easy to integrate their classical compute infrastructure in the cloud with the quantum service in the cloud.

And as a result, because they're used to that interface, because it integrates easily, and because they don't have to worry about some of the security measures, since that's already being taken care of by the cloud service provider, that tends to be their preference. So we see both. I know that doesn't help you with your answer.

Yuval: No, it's perfect. Do you see opportunities for cloud service providers, or even new data center providers, to use quantum as a magnet, to say, "We've got a great quantum computer in our data center, so now you should move some of your classical compute to us for ease of integration"?

Heather: I have heard some cloud service providers putting that out there as a kind of marketing technique.

I will say that's a new message I'm starting to hear. I think connecting the dots is, again, a fairly new conversation, but it definitely brings up food for thought, right? If you're going to be using a quantum system that's in the cloud, it makes sense to transfer some of that data into the cloud as well and start working with some of those other classical systems in the cloud, because of all the issues we just spoke about.

Yuval: You've been covering quantum for a number of years. What do you know today that you didn't know six months ago? What's new?

Heather: I try to keep up with quantum as much as possible, but every day, every week, there's a new announcement coming out from a new hardware vendor or a new software vendor. It's just an amazing time right now in quantum with all the advancements. So every day I'm learning something new. But one of the things we saw this past year, in 2022, was a bigger emphasis on post-quantum cryptography.

And I think the reason we're hearing more about post-quantum cryptography is that we're at this inflection point where quantum is moving from being an academic technology into commercialization. It's not ready for commercialization, but more and more executive-level individuals, as well as enterprises themselves, are really starting to take notice.

And so now they're starting to understand what quantum is and that it can solve these really difficult problems, and then making the connection that one of the most difficult problems it can solve has to do with security and how we encrypt our data at this point in time. So as a result of that transition, we have started to see more of an emphasis on and concern about post-quantum cryptography, and what enterprises should do now to start protecting their data, so that it won't get stolen now and decrypted later.

So that's one thing. Another emphasis I'm starting to see is on quantum networking and communication. Obviously, research has been done on that. This is not new to quantum, but with the conversations surrounding scaling, and using modular architecture to scale these machines in the future, comes the question of how you connect the machines.

What does that connectivity look like? That conversation is now starting to center more on quantum networking and quantum communication. So that's a new area that I'm excited to learn more about as we start hearing more about how these architectures will be developed and introduced.

Yuval: As we get close to the end of our conversation today, I'm wondering what IDC advises customers with regard to full-stack versus best-of-breed vendors. Is it best to go to a single vendor that has hardware and software and consulting and what have you? Or should an enterprise contract with multiple vendors…

And try to get them to integrate with each other?

Heather: Gosh, that's a difficult question. If there is a particular vendor that a company has been working with, where they get most of their classical compute, and they feel most comfortable working with that vendor, then perhaps that is the right option for them.

They feel comfortable; they've developed a relationship with a trusted partner. Perhaps the quantum specialists working for that vendor also have the subject knowledge needed to help develop these programs in combination with the technology. So maybe that is a better fit.

For companies that are still unsure about quantum, unsure about the type of technology and the types of problems, just starting out, it may be best to use a cloud service provider that gives them options for experimentation, to see what's out there, what they can play with, what works, what works better now, or to run short-term, small instances on these machines.

So I think it really depends on the comfort level. But if a company is unsure, IDC is definitely recommending taking a look at these cloud service providers and some of the quantum computing platform-as-a-service offerings that are out there.

Yuval: I want to ask you a hypothetical question. If you could have dinner with any of the quantum greats, dead or alive, who would that—

Heather: Okay.

Yuval: —person be?

Heather: Oh, goodness. You know, there's such concern about security. I think Peter Shor would be a great person to have dinner with, just to ask him, "What do you think? You have this algorithm out there, and now there's all this concern surrounding it. What are your thoughts?

Would you do it over again? How would you guide people?" That would be an interesting conversation, I think.

Yuval: Perfect. How can people get in touch with you to learn more about your work?

Heather: Sure. I'm on LinkedIn, if you want to reach me there: Heather West. I'm also on Twitter, and people can contact me through my email at IDC.

It's hwest@idc.com.

Yuval: Excellent. Heather, thank you so much for joining me today.

Heather: Well, thank you so much. This was a great opportunity. We'll have to do this again soon.