SAS Defines Hybrid Reality For Quantum Computing

0
SAS Defines Hybrid Reality For Quantum Computing

Quantum is huge. Because quantum computing allows us to step beyond the current limitations of digital systems, it paves the way for a new era of computing machines with previously unthinkable power. Without recounting another simplified explanation of how quantum gets its power at length, we can reference the double-slit experiment and perhaps the spinning coin explanation.

A coin sat on a desk is either heads or tails, rather like the 1s and 0s that express the on or off values in binary code. Quantum theorists would prefer we think of the coin above the desk, spinning in the air. In this state, the coin is both heads and tails at the same time. This is because, at the quantum level, both values exist until we make an observation of its state at any given point in time. We could further increase the number of positions possible (literally known as quantum superposition) by altering the angle of view we take on the coin, which is somewhat similar to how we work with qubits in quantum mechanics.

So then, Schrödinger’s cat is both alive and dead at the same time and the dummies guide to quantum entanglement is out there on the web if needed. What matters most now is how we will make practical use of quantum computing and where it will be applied for best advantage.

What Next For Quantum?

Across the universe of quantum entanglement, we’re at a point where quantum computing could redefine data analysis and model training in AI. This is the suggestion made by Bill Wisotsky, lead quantum architect and quantum computing researcher at SAS. In our near and immediate future, quantum computers could handle the complex calculations of AI algorithms much faster than classical computers and with less data, resulting in AI that can learn and adapt in ways we can’t currently imagine.

The main advancements Wisotsky and team are seeing inside SAS when speaking to customers includes work focused on fuelling the creation of a greater number of “good quality” qubits, which (in theory if not in practice) should get us to a point where we can tackle larger and more complex problems.

“Qubits are the smallest functional unit of QPU [quantam processing unit], analogous to classical bits. A larger number of computation qubits results in larger problems being solved on the QPU. Better quality qubits result in more reliable results that are less prone to ‘noise’ and instability,” said Wisotsky, in a company analysis blog.

As a quantum architect, Wisotsky explains quantum’s transformative potential to handle intricate computations and address detailed “combinatorial” analysis problems i.e. a mathematical term describing the number of elements that exist in a larger sum of values and the complexity of their arrangement. This is the type of analysis needed in all industries, but it has particular relevance for healthcare and life sciences and across finance and insurance.

Hope Lies In Hybrid

SAS underlines the fact that today, what has been most promising is the use of quantum-classical hybrid approaches, which aim at splitting computing processes and sending the pieces to quantum that quantum does best and the pieces to classical computing that classical does best. It’s slightly reminiscent of when cloud computing arrived and we were told that “cloud is for everyone, but not for every thing” – so therefore, similarly, it’s important to realize that not all problems will benefit from quantum computing.

“Problems need to be complex enough that classical systems struggle to solve them, only then will quantum properties be valuable,” explained Wisotsky. “For example, traditional computing might take days or months to run trial and error scenarios to find the best drug for a clinical trial, which impacts a hospital’s budget and personnel. Quantum computing offers the ability to search all solutions instantaneously, finding optimal drugs for trial much faster and more efficiently. By using both traditional computing and quantum computing in concert with one another, organizations can start to realize the benefits of quantum now.”

It is perhaps refreshing to hear enterprise tech firms talk realistically about the still-embryonic nature of an emerging technology in this way. With all the generative AI hype we’ve been working through for the last year or two, we need to handle this great power with really great responsibility. Extending and dovetailing with Wisotsky’s insight on this subject is Bryan Harris in his core capacity as CTO of SAS.

SAS R&D Factor

Harris talks about the quantum research work currently being carried out inside SAS R&D and, although the company is not building a quantum computer (it’s a discipline and skillset best left to specialized manufacturers and those tech behemoths with enough muscle to underpin development investment in this space such as IBM), it is actively exploring real-world applications and running actual use cases with third-party quantum computers.

In terms of practical tangible developments SAS researchers are investigating four opportunities for quantum computing:

  • Drug discovery: In the pharmaceutical industry, quantum computing is said to reduce the time and cost associated with discovering new drugs by simulating the behavior of molecules. This includes understanding the interactions between drugs and the complex biological systems they target.
  • Financial modeling: As the world becomes more digitally connected, the complexity of modeling risk is beginning to overwhelm classical methods. In the banking industry, quantum algorithms have the ability to improve the modeling of financial markets, portfolio optimization and systemic risk associated with hyperconnectivity.
  • Chemical simulations: Quantum computers can simulate the behavior of atoms and molecules at a quantum level with high accuracy. SAS researchers have noted that this is a challenge for classical computers because of the complex nature of quantum mechanics. In the science field, quantum computing can enable the discovery of new materials for improved sustainability to include much needed breakthroughs in batteries for electric vehicles.
  • Optimization: Fourthly here, quantum computing could dramatically reduce computational time by finding near-optimal solutions through its ability to search an entire space instantaneously, leading to cost savings in cloud computing operations and tackling previously unsolvable problems.

For an example of the above-noted process of optimization, in traditional cloud computing, running trial and error scenarios to find the best solution could take days or months, which is expensive. Quantum computing offers the ability to search the space instantaneously, finding optimal solutions much faster and more efficiently.

Quantum In Data & Analytics

With these four cornerstones under development then, what is the future for of this technology in data and analytics, especially given new post-quantum cryptography standards guidance from NIST? To put the question another way, what are some possible pros and cons resulting from quantum’s introduction to modern enterprises?

“With every new technology, there are opportunities and risks,” advised SAS CTO Harris, speaking to press and analysts this week in London. “First, quantum presents many positive opportunities for businesses to solve challenging problems that were previously unsolvable. However, quantum computers have the potential to break many of the cryptographic systems we rely on today for secure communications and data protection. This reality has put it on national agendas with the recent announcement of NIST’s post-quantum cryptography standards.”

As a result he says, there are two important workstreams that must happen simultaneously.

“First, organizations must allocate research dollars into quantum computing to understand how it be leveraged to create a competitive advantage in products and services. Second, IT and product development organizations must plan for the integration of quantum-resilient encryption algorithms to maintain the security integrity of their infrastructure,” insisted Harris.

Which Verticals Will Adopt Quantum?

As we have noted, not every business or life problem can take advantage of quantum computing. We need to look for application scenarios that must be complex enough that classical computing systems struggle to solve them. The sectors most likely to benefit include life sciences, banking and materials science.

The big question is, no… the “first” and most important question we might ask here is just exactly how will quantum computing change the data and analytics industry, or indeed other industries?

“The first major impact that quantum computing will have on data and analytics is search space optimization for AI,” said SAS’ Harris. “In many AI problems, the training of a model or machine learning requires the exploration of a highly-dimensional data space for potential solutions. With classical computers, searching this space can be slow and expensive, especially in the cloud. With quantum computing, the entire space can be searched simultaneously to find the best solution that can be used as a starting point in classical computing.”

Our takeaways here are, like quantum theory itself, both simple and complex.

Put simply, quantum computing is as occasionally fragile as it is magnificently powerful and we’re still at a prototyping analysis stage with this technology, but we’re quickly coming out of that phase into real world applications. Put in slightly more complex terms, our quantum reality is likely to be a hybrid combination of traditional compute architectures made up of CPUs and GPU as also we now add QPUs into the mix.

As for Schrödinger’s cat, the lid isn’t off that box yet, thankfully.

link

Leave a Reply

Your email address will not be published. Required fields are marked *