IBM Quantum Industry Webinar Series: Quantum Computing for the Automotive Industry
Hello and welcome to the IBM Quantum Industry Webinar on Toward quantum utility in the automotive industry. Today we are joined by Dr. Gabriele Compostella, who is a quantum industry applications consultant for the automotive industry at IBM. He supports companies in their journey toward quantum utility and relevant industrial solutions. So he'll be taking us through the work that we are doing with clients in this industry, of course, with applications ranging across other industries as well. And so we'll look forward to hearing from him. A few notes.
We are recording, so we'll be sharing a recording with you in the coming days. And we also encourage you to ask questions in the Q&A dialog window throughout the session. And Dr. Compostella will be addressing those questions as many as we can at the end of the webinar. So thank you for joining us today. And I will turn it over to Dr.
Compostella. Thank you Jennifer, thank you very much. And thank you all for being here. So the goal for this webinar today is to give you an overview of where we see potential for quantum computing to provide business value for the automotive industry. So let me get started.
Many people think that quantum computing is just a new technology, but actually we think that quantum computing is the second quantum revolution in the history. And, and what was the first? Well, the first quantum revolution happened 100 years ago. This year is the 100th year since the formulation of quantum mechanics. And basically the realization back then was that nature actually is discrete or comes in quanta.
And so this means that energy, velocity, light position, etc., they are all quantized at the fundamental limit. So this had a huge impact on technology development and enabled applications like lasers and MRI's, solar cells, transistors. Some of them you see them here on the slide. And it's hard to define how much value this revolution produced, but it's estimated to have created tens of trillions of dollars worth of impact. Well, we expect that quantum computing will have a similar impact.
This second quantum revolution has to do with realizing that nature works and computes beyond simple arithmetics. And in fact, this is bringing together computer science and the physical sciences, together the fundamental limits of computing. So you can think of it as quantum mechanics, giving us new ways to solve problems that are actually intractable on standard classical computers.
Well, but, what are these problems, actually? One class of these problems is, has to do with molecules, modeling molecules, atoms, electrons and quarks, with unprecedented accuracy. You can think of applications like, developing, better batteries for vehicles or energy storage systems, designing lighter, stronger materials for airplanes, discovering new classes of antibiotics to counter drug-resistant bacterial strains, or, designing optimal superconductors for MRI's or for electromobility. But moreover, there are other applications. Another class of problems has to do with the solution of complex algebra in exponentially large spaces, or with finding hidden patterns in structured problems. So the most famous application in this area is breaking asymmetric cryptography with the Shor's algorithm.
But here you can also think of other applications that we have in this slide, like improving anomaly detection or the detection of rare events like, I don't know, financial frauds, improving patient outcomes by designing optimal cell-sensitive therapeutics, or improving risk management with better time series forecasting, or even optimizing vehicle routing for large scale logistics. And, what's interesting is that given the promises of this second quantum revolution, we are observing that investment in quantum computing is accelerating at an unprecedented pace. So the World Economic Forum reports $40 billion of global investment. The IDC reports that customer spending is growing at about 50% compound annual growth rate. And according to BCG, in 97% of the companies are planning to drive investing in quantum are planning to either increase or maintain their current investment in this technology. Well, all of these investments are driven by the estimation that given these many potential applications, the value created by quantum computing at technology maturity could be very high.
And for example, here, there's some data from BCG that predicts over $500 billion in value created by quantum computing. And you also see how different industries will be affected according to their prediction and where most of the value could come from. So the impact on the automotive industry is expected to be significant. So let's take a deeper dive.
And let me give you some more details of where we think there might be valuable use cases to address with quantum computing in the automotive sector. So, we believe that quantum computing will extend our, capabilities to solve complex problems in automotive that may be beyond the capabilities of current classical hardware. And here's a list. As discussed, some of these problems are in the area of chemical and materials simulations. And, you know, in automotive here, you can think of use cases related to battery development, materials development, for example to lightweight cars, or simulation of chemical processes, calculation of corrosion reaction rates. We do also see potential for applications in quantum machine learning.
So here you can think of applications around quality control, fault detection, predictive maintenance, demand forecasting and many others. We expect quantum computers to also contribute to optimization in such use cases. And here we have a problem reduction scheduling optimization, supply chain optimization, vehicle routing optimization, or other complex optimization problems that are relevant for shared mobility or vehicle to grid scheduling.
And in the longer term, we expect quantum computers to provide new ways to solve complex linear algebra problems more efficiently than what we can do right now. And these problems are at the basis of, engineering simulations like finite element simulations or computational fluid dynamics simulations that are important in automotive design. And there is also the idea that potentially there could be applications in, model training for autonomous driving, though these use cases are still very aspirational. So why, you know, quantum computing in automotive? So the reason we should look at new ways to solve those problems that we've just listed is that the automotive industry is transforming and is experiencing new challenges because of these transformations.
So challenges related to, for example, manufacturing complexity with automakers having to adapt their production lines to handle a wider variety of models, a wider variety of options and configurations. There are challenges due to electrification, with the industry rapidly shifting towards electric vehicles and with a projected market share of up to 70% by 2040. The industry is also transforming towards shared mobility. There is a growing consumer trend to that is focusing on car sharing. And this is basically changing the concept of car ownership.
And finally, another transformation is towards autonomous and connected vehicles, with companies investing heavily in software innovation in order to integrate AI, sensors and connectivity into their cars. All of these transformations are interconnected, and together they are basically reshaping the automotive industry and required manufacturers to adapt their strategies to invest in new technologies, to reimagine their role in the future of mobility. So I'd like to see together with you more in detail how quantum computing could support automakers in this transformation. So if we look at manufacturing complexity, we observe, again, increasing complexity in manufacturing with technical equipment and machinery that has been increasing much more quickly than the volume of cars produced in the last years. We also see manufacturers facing rising costs of almost 10%, compared to 2% to 3% of the previous years. And there is a need for more flexible production and more adaptable processes in order to increase competitiveness. Here,
quantum computing could complement classical solutions, could help, for example, improve OEE, the overall equipment effectiveness. So for example, this could increase, for example, quantum algorithms could improve performance right by optimally scheduling jobs, or optimally scheduling shifts in a plant. They could be used to improve quality by having better models for early detection of defects. And so this could reduce 'scrap and rework' and reduce costs. Or they could improve availability by creating better predictive maintenance tools that reduce unexpected downtimes and breakdowns.
And just to give you an example of how quantum could contribute to this area, we talked about optimal shift scheduling in the previous slide. Here's a problem that IBM has explored together with Woodside Energy. So the challenge for Woodside was that, their maintenance workers are deployed inefficiently with up to 65% downtime historically, and finding the optimal schedules to dispatch workers with the right skills to the right maintenance tasks can quickly overwhelm classical methods, especially when the number of decision variables and the constraints that have to be taken into account are large.
So in this case, we implemented a quantum optimization algorithm on our hardware, using some representative synthetic data. And you can see some sample results on the right of the slide. So the main point here that I want to make is that we could show that the scaling properties of this algorithm are better than classical equivalents. So we expect that when quantum hardware will mature, it could potentially outperform classical methods in terms of solution quality. So effectively in this case reducing the inefficiencies and the costs that are, that are incumbent for maintenance. And maybe algorithms like these could also be using other forms of optimizations for, for plans or for production schedules.
If we look instead at the challenge of electrification, we observe here different aspects. So first of all, high capital investments with spending on electric vehicle batteries alone expected to reach $400 billion by 2030. At the same time, we see that customers have very high expectations. For example, in Europe, customers require, want to have at least 500km driving range in order to even consider switching to electric vehicles. And moreover, there are concerns about safety, charging time, costs, that represent additional barriers for adoption of electric vehicles.
Here, there could be the potential to apply quantum algorithms for chemistry. And they could offer new ways to improve the precision of the simulation of the properties of the materials that are used for batteries. And maybe they could help extend what can be explored computationally in this area. So we can think of, this approach helping us improve, for example, range by studying compositions with, with higher densities or reduced costs of batteries, by discovering cheaper materials computationally, or contribute to, increasing the safety of batteries by exploring the properties of next generation batteries, for example, calculating properties of solid state batteries. And here an example of of of using these algorithms for battery chemistry comes from our collaboration with Mitsubishi Chemical. In this work we explored the chemistry of lithium oxygen lithium air batteries.
And these batteries are considered a next generation energy storage technology because they offer energy densities that are higher than lithium ion batteries and that are comparable to gasoline. You can see this exemplified in the plot on the left. And this, this, this large energy density makes these type of batteries particularly attractive for application in electric mobility, since they could provide longer driving ranges. And additionally, since they use oxygen from the air as a, as cathode material, these batteries can be much lighter. And they could even come at a lower cost since they do not need to, scarce scarce materials or rare earths that are difficult to, to supply.
So however, there is a challenge in, in these, these batteries, that is that both the charge and discharge process in these batteries are complicated. They are very sensitive to the surrounding environment. This has somehow limited the progress of the technology. So IBM and Mitsubishi Chemical here invested, investigated, sorry, together, how we could, simulate with quantum algorithms the rearrangement of the lithium superoxide dimer, which is, one of the components of this reaction. And the, this element requires an accurate description of bond breaking and, and the bond formation through transition state. So the transition state that you can see in the, in the right of the, of the slide.
And in the plot in the bottom right, you can see the results of the calculations that we performed of the energy of the reactant of the transition and product state for this reaction, performed with two different IBM quantum hardware. And you can also see that there is a very good agreement with the exact solution using full configuration interaction. These are the three lines that you see on the very bottom of the plot. And in this full configuration interaction for this particular case can still be calculated classically. But the point here is that, these results give us confidence in the method that we implemented. And we expect that, with hardware maturing, we will be able to explore more complicated chemical problems.
Moreover, we used similar methods together with Bosch to develop a hybrid quantum classical simulation framework that we used to calculate the electronic structure of a cuprate superconductor. In the plot here, you see the results that we obtained with this workflow and how they correctly reproduce the single particle spectrum from spectroscopy, from, from experimental spectroscopy for this material. So again, so more, some more confidence in these methods that they are able to replicate the real properties of these materials. And more recently we developed a new algorithm for chemistry applications on, near-term quantum quantum computers.
And this method is called sample based quantum diagonalization, SQD. And it allowed us to achieve the largest and most accurate quantum computations of chemistry to date. We have performed three sets of electronic structure experiments using the Fugaku supercomputer hosted at RIKEN to assist an IBM quantum processor. In the three panels, you see the results of these three sets of experiments right, at increasing molecule complexity. So first you see the, study of the nitrogen two bond breaking, and then the energies of two different iron sulfur clusters that are very difficult to model with classical methods, since they are strongly correlated systems.
So this work shows that, even with a noisy quantum computer, we can perform already, precise calculations of, many body physics, and and this method has been used already, has been already applied to, different industry use cases. Here you see some examples from internal work that we did at IBM, examples from our collaborations, with Cleveland Clinic and examples from our collaboration with, with Lockheed Martin. So if you're interested, I invite you to to take a look at the papers for more details. But the message that I want to give here is that the success of these SQD methods, with these different applications, is increasing our confidence that this approach, with these approach, quantum computers will be able to solve accurately chemistry problems that are relevant for battery development in automotive. I also want to point out that there are many other use cases beyond, let's say, pure chemistry that are related to electrification, where quantum computing may provide some forms of benefit.
You see, at least in this slide, divided by problem class, spanning from design to manufacturing and testing and to usage of the batteries on the road, use cases here include quality control, battery lifetime predictions, different forms of optimization, different types of modeling of complex processes that are relevant for batteries and without entering too much into the details, I just would like to show you one particular example that is related to, to smart charging. Here, in this work, IBM and E.ON studied together how to leverage electric vehicles as energy storage systems, to provide additional capacity to the electric grid. So here the problem of deciding when to charge and when to discharge vehicles optimally to support the grid is extremely challenging for classical methods.
And in this work, we first focused on learning optimal policies from data. So we expect that quantum methods will extend our possibilities to solve similar complex problems in the future. And they might be used, for example, to guide dynamic pricing algorithms that, that could incentivize users incentivize, sorry, users through participating vehicles to create, the vehicle-to-grid system at optimal times at optimal locations.
And this would benefit both users and operators potentially reducing costs. You know, we are we are trying to use already algorithms. If we take a look at shared mobility, there are some observations here as well. So first of all, consumers are changing habits and customers are more concerned about sustainability and are considering to even to replace their private vehicles. At the same time, this could represent a new business opportunity for car makers, with an estimated market of up to $1 trillion in 2030, by 2030. This opportunity, however, also comes with new challenges that are related to efficiently and timely dispatching vehicles to on demand services so that they could be profitable.
And this often requires the solution of very large and difficult optimization problems. And, and here quantum computing could help. Right. So with optimization problems that cannot be solved exactly by classical solvers, quantum algorithms have the potential to identify more optimal solutions, so, solutions that have a lower optimality gap or identify more diverse solutions with respect to classical algorithms.
Or potentially they could offer quadratic speedups for some particular implementations of these problems. And in general, the result would be that this could save, this could help save energy, reduce traffic, increase customer satisfaction, and basically improve the overall revenue from these types of services. And we've reviewed with the Hyundai Motor Company, we explored a use case in exactly this area. So we focused on vehicle charging, routing and scheduling optimization with multiple objectives, which is a fundamental problem for both logistics and shared mobility. We analyzed a fleet of electric trucks that deliver goods to customer locations. You see them in the plot.
We indicated with, with the c's. And along the way, these electric vehicles can stop to recharge at intermediate stations that you see labeled by, by s. So the objective of the optimization was that given locations, given charging stations, given the fleet of vehicles, given some time windows and capacity restrictions, we wanted to minimize the distance traveled by the fleet, minimize the time to deliver these goods, and also minimize the number of trucks in the fleet.
This is a NP hard problem, so it's a problem that to be solved requires computational resources that increase exponentially with problem size. So, and it's a problem that is particularly challenging for classical algorithms. And in the table on the right, you can see, how this problem becomes increasingly challenging to solve up to a point that with more than 200 locations, even state of the art classical solvers like CPLEX struggled to find the optimal solution and return an approximation that has some gap from optimality, and from the business perspective, a gap from optimality means that, so that you get solutions that where the distance traveled or the number of vehicles in the fleet may not be optimal. So a solution that is more costly than optimal or where customers have to wait more, for their deliveries. Well, together, we developed a sampling based variational quantum algorithm that can be used to determine the Pareto front of this problem.
So the Pareto front is basically the set of optimal trade offs between the different optimization targets. Right. In the plot on the right, you see some points on the Pareto surface that represent the trade offs between distance and time.
And this algorithm represents a first quantum approach for multi-objective optimization on a noisy hardware. And the interesting thing is that it showed some favorable scaling properties. So we expect that with similar approaches, we might be able to solve increasingly large instances of this problem as hardware matures and probably find, some benefits in applying quantum computing, computing here. Moving on to, to the challenges related to autonomous and connected, connected vehicles.
So here we observe a large business opportunity for automakers that could differentiate their products and also access a market of up to $400 billion in revenue by, by 2035. At the same time, we also see that there is a requirement for high capital investment here, with companies having to invest in tens of thousands of GPUs in order to train and test their autonomous driving systems. This, this business opportunity also comes with, with new challenges, on safety, regulations, reliability.
And this will require companies to iterate very quickly on their software and develop new capabilities. Here some of these applications are more long term, but here we expect that research on quantum machine learning and quantum optimization could one day unlock our ability to optimize, for example, when and how over-the-air updates are distributed to vehicles, maybe based on customer preferences, based on network conditions and so on, or could enable, better software verification and testing with respect to what can be done right now. Additionally, there are some ideas on using quantum models to generate more representative data sets, synthetic data sets, for safety critical scenarios. And this could reduce cost of data acquisition. And finally, we hope that in the future, quantum algorithms may be used to accelerate the training of large AI models that are used in autonomous driving systems.
Though we expect this to be a very long term prospect. So, summarizing somehow, I hope I could convince you that quantum computing applications can play a role in supporting transformation of the automotive industry. Quantum algorithms offers several opportunities for process optimization that could be relevant to manage manufacturing complexity.
In the field of electrification, quantum algorithms may allow for more accurate calculations of important properties of battery materials, and also could be used for many other use cases, as we've seen. For shared mobility, quantum optimization may improve solutions quality and diversity and could help reduce costs or increase customer satisfaction for these services. And finally, quantum computing may also support different use cases related to autonomous and connected cars and contribute to the safety of autonomous systems. So, but where do we stand right now in terms of technology development? There have been several breakthroughs, that brought us into what we now called the era of quantum utility. Moving from left to right.
So in June 2023, we demonstrated for the first time in history that quantum computers could run circuits beyond the reach of brute force classical simulations with reliable results. And this made the cover of nature back then. In December 2023, we announced another breakthrough, this time in the quality of our quantum processors with the new gate architecture of our Heron processor that is offering, up to a five times improvement in error reduction with respect to the previous generation. And in March 2024, the third column, we published the paper that laid the theoretical groundwork fornpractical error correction. And this paper detailed a new quantum error correcting code or an error correcting algorithm that we created that is about ten times more efficient than the previous known ones. And that brings error correction closer, making it, scalable, within reach.
And all these scientific advancements do inform, and do guide, our IBM quantum development roadmap that you can see here. And that basically represents our long term commitment towards our users. On the roadmap, you can see that we have designated, 2029, on the right with the Starling processor, as the date when we will realize quantum error correction and we will be able there to run 100 millions of gates on, on 200 qubits.
So at the same time, while we are developing our hardware, we are also implementing different tools to make it, to make our hardware easier to use for our users. And here I just would like to call out, one of the recent developments in this area, the introduction of Qiskit Functions. And you can think of the Qiskit Functions catalog as a sort of, an app store, for quantum computing applications, where you will find, a catalog of services that could accelerate research, either built by IBM or by third party partners. Currently, we have applications, applications functions in the areas of chemistry, optimization, and machine learning. And also we have circuit functions that, that are thought to accelerate lower level calculations that are related to transpilation, error suppression, error mitigation, and other functions like these.
So the idea of these functions is to abstract away some parts of the quantum software development workflow so that they could simplify and accelerate, algorithm discovery, algorithm development for our users. And, and here are some, some, some notes about our timeline for the next four years and what you can expect. In 2025, in this year, we will focus on demonstrating how to integrate quantum and high performance computers in a quantum-centric data center. In 2026
our target will be to demonstrate quantum advantage. So to identify a problem that a quantum computer will be able to provably solve beyond the capabilities of classical solvers, sorry, classical computers, meaning either finding better solutions, finding them faster, or in a more cost effective way than by using classical computers alone. And then in the future, in 2027, we aim at demonstrating a viable path for error correction and to implement it, in the first ever corrected quantum computer by 2028. And given this, why you why should you become quantum ready now? Right. So we believe it's worth starting to become quantum ready already now, as it takes time to prepare organizations for this new for this new, technology, for this new computational model. In particular, organizations will need new capabilities to create and validate new algorithms, new data models, new workflows that are required to solve applications to solve, sorry, use cases in their domains.
There is a huge gap, there's a huge gap in near-term talent, in this, in this point, which makes it time consuming to hire and train a quantum workforce. And currently, according to McKinsey, there is just one qualified candidate for every three job postings in quantum computing. So it's very difficult to find the right people. And finally, also securing hardware access will become even more critical when quantum achieves, commercial advantage over a classical solution, since the demand for access will potentially will likely, overwhelm the supply. So
I'd like to conclude with some recommendations for our industry leaders. We have seen that quantum computing has a high potential for disruption, right. So we suggest to establish a strategic plan around this technology that takes into account some of the things we discussed today. So it takes into account the arrival of quantum error correction in 2029, the expected supply demand mismatch or the talent shortage that we just discussed.
We also suggest to evaluate opportunities offered by quantum computing, through proof of concept applications. So these are so that you can identify interesting applications for your domain, understand how to implement them, develop quantum capabilities and talent that could be useful for these applications. We also suggest to join an ecosystem to benefit from the experiences of others. All right. Since nobody can do this entirely alone and gain access to hardware, gain access to open software development platforms, like, for example, our Qiskit. And finally, we, as we have seen, quantum utility is already here, industry leaders are building advanced quantum capabilities, and IP is being secured right now.
So we also saw that the learning curve is very steep. So companies may require years to master the technology. So we our recommendation is to start already today. And of course we at IBM here are happy to support you if you're interested in collaborating with us.
So if you want to know more, I invite you to take a look at the links that are available in the platform or, reach out if you want to hear more about how quantum computing could be useful for your business. And with that, I'd like to thank you very much for your attention. If you have any questions, feel free to post it in the chat and I'll do my best to answer. Thank you, thank you Gabriele. We really appreciate that talk.
And yes, as he mentioned, please enter your questions in the Q&A box on your console and we will try to get to as many of these as we can. I see one, that I can mention here. I am teaching a pre-college class, on unlocking the potential of quantum computing, real world use cases. Can I use some of these resources? We have many great resources that would be fabulous to use in a course like that.
And there are a couple of links on the console that you can see, to IBM Quantum Learning, for instance, where there are, many, many great, free resources that work really well, in an undergraduate class setting. And also if you're looking for, additional real world examples, I would encourage you to go to the IBM Quantum Case studies, which have videos and written content on many of these examples in, in various industries. That would be a good resource there. All right. Let's look at some of these questions here.
I think one clarification question. Gabriele, the Hyundai EV charging problem was this implemented as a quantum circuit or was an annealing approach used? No. So this was, this was implemented not as an annealing approach. So our quantum computers are gate based quantum computers.
It was a variational algorithm based on, sampling, on a sampling approach where we were sampling solutions and, that could, let's say, solve the problem, after having split it into two smaller components, but so no annealing, pure gate based with a variational approach. Okay. Thank you. Yes.
And a follow up question on that. How is gate based quantum computing, compared to quantum annealing ones at solving optimization problems? Maybe you can give some examples of. Exactly.
Yeah, I think I think this is a difficult one in the sense that many people in the community have different perspectives on what are the pros and cons of the two technologies. So let me just focus on, on our own technology. Gate based model quantum computing allows you to, let's say, to implement any type of circuit that your, that you, that you may want to do.
So it allows you to implement applications that go beyond optimization problems. In comparison to annealing, which is much more suitable for, for optimization problems only. So this is, this is it. So regarding instead the quality of solutions, it's difficult to give, let's say a blanket statement on this. It strongly depends on the use case, but not only on that. It also depends on how the problem is mathematically formulated, how it is represented.
What kind of choices people use to base decisions, and so on. So, I find it hard to give a comprehensive answer here. It depends on the case by case. And that's a great, great insight there.
Thank you. So a couple of questions about computational fluid dynamics. Which, you know, people acknowledge is important for manufacturing and designs for automotive. Can you speak to the benefits and challenges, with quantum computing in this, in this area? Sure. Yeah, sure. It's a very important topic, and I know that many people that are interested in potential applications of quantum computing to these engineering challenges, which are pretty fundamental in any type of manufacturing, manufacturing engineering problem. So here the idea why we call out potential applications of quantum computing to computational fluid dynamics is that, there are already some algorithms that are known, in the community.
So theoretically at least one of them is the HHL algorithm that are expected to be able to solve quantum, sorry, to solve linear systems or linear algebra systems, better with a quantum computer than what can be done with classical computers. When I talk about better, I mean with better scaling properties, right? So the best classical algorithm to solve a linear system scales with n, where n is the number of elements, basically, you want to solve for, while this HHL algorithm scales with log(n). So it's it has a better scaling property. So this means that, everything else being equal, there is a possibility of using these algorithms on quantum computers to either address larger simulations or being faster in addressing simulations that we that we that we address already with classical computers. However, these algorithms do require error correction and do require large quantum computers.
So it's something that we see as something that could be potentially feasible in the future to start looking at these kind of applications. But it's something that we we are not able to implement right now. But we think still that this could be an interesting application long-term for quantum computers in problems that are very, very relevant for the manufacturing industry. And I hope this answers the question. Yes. Thank you.
Okay. And along similar lines, a great question about thinking, you know, thinking about the work being done today on these proofs of concept and how we imagine then scaling, you know, as we approach, quantum advantage here. So a question, the question is, how can a proof of concept provide value to organizations today? Can these quantum proofs of concept now be applied to future quantum advantage hardware? Yeah, I think so.
So I think it's that well, first of all, I think it's a very good point. And I, I'm totally I mean, of course it depends on the point of view here, but I'm totally convinced that, these kind of proof of concepts that we do right now, are extremely valuable in realizing quantum advantage. It will be valuable to, to realize future applications.
So first of all, there is value that you extract immediately by running these proofs of concepts. That is, well, upskilling teams, learning how to use quantum computers, learning how to program them, learning how to reformulate your computational problems, in a way that is suitable for quantum hardware. We also have to remember that programing a quantum computer is not as simple as learning a new programing language. There is a bit more to it. We also need to formulate our problems in a different way to, to match with this, with this new paradigm of computing. So this is a learning process that needs to be done.
And proof of concepts offer a way to, to do this. And at the same time, the kind of code that you develop is going to be useful also, as the, as the, as the hardware matures. Right. So this is our, our promise with Qiskit, right?
That what you, what you develop is going to be, reusable, with, with our technology maturing. Right, that Qiskit will be compatible with our future hardware. There are also other things that could be valuable.
I mean, you, by engaging in proof of concept, you also do all the work that is required to understand where actually quantum computing could provide benefits in your organization. You can see where are all the opportunities, where the use cases are that are more valuable to you and where quantum computing may provide some may maybe not now, but in the future, potential for, for benefits or for return on investment and proof of concept, also allows you to calculate when return on investment might be realizable according to how the technology will mature. What is that? What are the requirements? Yeah. And the other thing that is often I think, dismissed, but I think it was, at least in my previous experience, kind of, useful is that, often when you implement proof of concepts of quantum computing problems, you have to also dig deeper in what is the state of the art of classical algorithms applied to, to your particular problem under scrutiny? And often it turns out that, while you do quantum computing, you also discover new things about your classical problems that might help you already right now, improve your solutions to it, but at the same time, build future proof solutions that could then run on the quantum computer when hardware, when the hardware is mature.
So I think it makes sense to engage in proof of concept. And there's a lot of value that could be extracted from them already, right now. Yeah, absolutely. And we have a number of questions about quantum and HPC. Of course, you know, we have the statement that we expect to see quantum advantage in the next two years if the quantum and HPC communities can, can work together. And you shared the example of sample based quantum diagonalization methods.
So we have a couple of questions in the chat. For people looking to learn more about these methods, can you point them to, resources or examples in which, you know, those techniques are currently being used to extract utility from quantum computers, in the near term? So I think I called out, a few publications that, are available online. Maybe we'll find ways to, to share the links. These are, these are ways to catch up with what we did with these different partners and what we did internally. To, by using SQD on particular problems, I think we have some tutorials as well that explain how to use it, practically, on Qiskit, and, I don't want to be, to be to, to, to, to make a mistake.
So please correct me. But I think we also have some Qiskit add ons. So these are basically components, of reusable reusable pieces of code that, that, that our users can implement with Qiskit in order to solve particular problems.
Well, we have a few add ons there to simplify the implementation of SQD for new, for new applications. And that make the experimentation with this method much faster. So you don't have to implement it from scratch yourself reading the paper, reading the papers. So I think there's a there's a lot of, of potential resources that you could use to learn more both about the method and try using it for, for, for applications that are interesting for you.
Yeah. There's a lot of great research being done in this area. As Gabriele called out, there is an SQD add on, which you can find, you know, it's available for everyone to use. It's a modular you know, research tool that can help you scale your workloads to, you know, the utility scale and, you know, just make, you can really accelerate your research in that area.
And there's also several chemistry functions as well, like the QunaSys chemistry function that can be useful to explore as well. So I would encourage you to check out our documentation and guides. Like Gabriele mentioned, there's one specifically for SQD where you can get started working with that today.
And I think we'll have to pick one last question here and we'll try, you know, we can try to follow up with any additional questions. Let's see. Okay. And maybe just building on the Qiskit Functions one. So you know, you mentioned Qiskit Functions. What tools are available for users willing to develop their applications? These are, as Gabriele mentioned, a great opportunity for, people to, get started with, you know, running workflows, quantum workflows without knowing everything that's going on under the hood. So, what would you recommend? Well, there are many others.
So Qiskit functions are one way. So probably the higher level libraries that you can look at. We recently released the Qiskit, version two.
So our framework for quantum software development, it's one of the most popular frameworks in the industry. It's selected by, I think, more than 70% of the quantum quantum algorithm developers. Also, it's also compatible with with other frameworks. So you you can use it as sort of lingua franca, for, for quantum computing development. So this is one aspect, you could you could start learning Qiskit and you take it and let's say from the, from the lower programing level and try to build your applications.
As we have mentioned, we have Qiskit add ons. So some components that can be reused in order to simplify parts of, of the, of the implementation of, applications according to the patterns that we see for quantum application development. And, and then again, we have the functions growing up to the, to, to the stack. And I think we have also other libraries that are, that are supported by our working groups. So we have libraries for chemistry, libraries for optimization, that are supported by our working groups.
So, groups of partners that work together with us on particular problems, on advancing quantum computing, on particular problems. And these are also available as examples that you could take inspiration from in order to develop your own applications. But I would invite you all to take a look at our learning web pages. Read the quantum platform. There's really a lot of content that can guide you through all your journey and learning more about quantum computing from the theoretical aspect up to those more connected to programming. And I think we there's a lot of great content that our colleagues have created.
And I'm sure this could be very useful, independent of what kind of journey you want to take in this technology. Great. Thank you. Yep.
So hopefully, you know, you all have enjoyed learning more about our roadmap and the work we're doing to demonstrate you know, business value with quantum computers. We're excited about our work with our clients and partners. And, you know, welcome you to reach out, connect with us on the IBM Quantum LinkedIn. You know, look at the IBM Quantum Platform, IBM Quantum Learning that we've referenced, and get started today. And after this webinar closes, we ask that you share your feedback.
We appreciate any recommendations you have for content. You know, tell us what's useful to you. The survey will ask you a variety of questions. So please do fill that out. When this webinar closes and there'll also be opportunities to reach out to learn more about how you can get started getting quantum ready today. Thank you everyone. Thank you.
2025-04-30 07:03
