Hope And Hype Of Quantum Advantage

In the past ten years, quantum computing research has made progress. Now, we can distinguish between what's real and just talk. Microsoft thinks everyone needs to work together to get the most out of quantum computing. Our aim is to assist and concentrate this genius where it can really make an impact. We teamed up with Torsten Höfler, a specialist in high-performance computing, to create a guide for quantum application development called "Disentangling Hype from Practicality: On Realistically Achieving Quantum Advantage." It's just been published.

Quantum computing has the potential to solve tough problems but not all of them. People are exploring various applications for quantum computing such as logistics, finance, and biochemistry. Everyone is hopeful but we need to be realistic. Quantum will have the most impact on certain areas. Quantum physics governs which problems can benefit from quantum systems. It's important to find algorithms that justify the cost of these systems. Problems can't rely on huge amounts of data as it's difficult to get data in and out of quantum systems.

Quantum and classical systems differ in their traits, and quantum computers are applicable to a distinct set of issues. Achieving quantum practicality is difficult due to the unique features that enable quantum computers to function effectively.

Quantum Usability Framework

Which problems will benefit the most from quantum computers is something we need to figure out.

We made a model to see what a quantum computer can do. We think a quantum computer with 10,000 logical qubits will be made soon. This computer will have one million physical qubits. We compared this to a single state-of-the-art GPU in a classical computer.

Quantum systems are better as they use nature's quantum foundations. This means they do better than classical computers for certain problems. But, quantum computers need more time for operations. Therefore, quantum algorithms must be faster than classical ones to make up for the longer time. But, what does this mean specifically?

We chose two weeks as our break-even point for analysis. A quantum computer must perform better than a classical computer on problems that take two weeks or less to solve. If we compare a hypothetical quantum computer to a single classical GPU, it needs to be more than quadratically faster to be useful. Super-polynomial speedup is what we ideally want. This is a significant discovery as many quantum computing applications rely on the quadratic speedup of certain algorithms like Grover’s algorithm.

Quantum computers don't have much bandwidth to run operations because they're complex. Even a bigger quantum computer can only handle 1/10,000th of the bandwidth of a specialized computer processor. Graphics processing units are often used for machine learning. Quantum computers can only process limited data sizes, so they can only solve simple versions of problems.

To know what quantum systems can do now and in the future, we need special algorithms and problems with less data. If we use these for different areas, we can see where quantum computers will be most effective later.

Quantum Computing Can Drastically Aid Computational Chemistry And Material Science

Quantum computing is great for simulating chemicals and materials on a quantum level.

The world has lots of problems connected with chemistry and material science. For example, better electric cars need better batteries. More powerful cancer medicine needs biochemistry knowledge. Also, we need materials that last and then break down quickly. All of these things require research into chemistry and material science.

Quantum computers are good for chemistry and material science. The Stone Age, Bronze Age, Iron Age, and Silicon Age are all named after materials. Chemistry and material science are important. They impact 96% of manufactured goods, which impact everyone.

We found out that quantum computers are great for chemistry and material science problems. They can quickly simulate the interactions of one chemical. This is because there are only a limited number of electron interactions to represent. On normal computers, this is really hard to do, but on quantum computers, it's easy and fast. This falls under our guidelines for using quantum computers.

To solve hardest chemistry and materials science problems, we need quantum computing. But, we can make progress with Azure high-performance computing. Johnson Matthey and Microsoft Azure Quantum chemists have used this to speed up quantum chemistry calculations for hydrogen fuel cell catalysts. By combining high-performance computing and specific quantum functions, they reduced the workload time from six months to one week.

"Innovation: Where To Focus"

We found out that classical computing is better for apps with big datasets. Quantum systems don't have enough bandwidth. They can't do stuff like searching databases or training machine learning models on big datasets. Examples are drug design and protein folding, which use Grover's algorithm. Also, weather and climate prediction, which use large sets of equations.

Quantum computers have limits for data input and output. This means they can only solve smaller problems. People building such applications will find best results with Azure HPC and AI services. If focusing on quantum research beyond materials or chemical science, Azure Quantum supports your exploration with tools like Resource Estimator and cutting-edge hardware. We are keen to learn what you discover.

I really want to see quantum's benefits happen while I'm alive. It's what drives me. But we need to work together to make real progress. Scientists need to focus on the areas with the most promise in order to see quantum benefits on a large scale. So, I urge this community to focus on learning and researching computational chemistry and material science applications.

Boost Scientific Advancements With Azure Quantum

Come join us if you want to speed up your chemistry and materials science R&D. Check out the latest edition of the Microsoft Quantum Innovator Series to know more about how we can solve quantum problems today and in the future.

Do you want to meet our chemists and quantum architects? They can teach you how to accelerate your research and development. If you're interested, contact the Azure Quantum team at [email protected].