Berkeley Lab Amplifies Power and Usability of Quantum Computing

The difficulty of sending a person to Mars or another planet is less about the brute force needed to send a rocket that distance and more about the myriad complexities of the mission.

The same is true when we go from the galactic to the molecular: simulating certain quantum-level phenomena could be hugely impactful, but our current tools aren’t up to the job. Today's supercomputers don’t lack for processing power, but they are not built to simulate the physical complexity of the scientific problem. 

The key that unlocks a slew of findings for physics and chemistry researchers maybe quantum computing. 

For Bert de Jong, a researcher at Lawrence Berkeley National Laboratory (Berkeley Lab) and a computational chemist by trade, the boundaries of classical computing for investigating increasingly large, complex chemistry problems led him to begin exploring the potential of quantum computing. 

"Exascale computing, quantum computing, and machine learning to me are tools in my toolkit to tackle the science problems in chemistry that I am interested in, such as carbon capture, water desalination, and renewable energy technologies,” de Jong said. 

de Jong believes the anticipated power of quantum machines could open the door to a wide range of research achievements in chemistry and beyond, and his involvement in a number of Lab- and DOE-based projects reflect this. In Berkeley Lab’s Computational Research Division, de Jong directs the multi-institutional program known as AIDE-QC, which stands for Advancing Integrated Development Environments for Quantum Computing through Fundamental Research. In addition, he is involved in quantum efforts in Basic Energy Sciences and High-Energy Physics and participates in Berkeley Lab's Advanced Quantum Testbed and Quantum Systems Accelerator research programs. 

“When it comes to quantum computing, working together with the hardware developers and science domains advances the software and algorithms to make it a useful tool for scientific discovery," he said.

An Integrated Research Effort

Much of the attention on quantum computing has so far focused on the hardware, which is still proving its efficacy and consistency. But software, where de Jong’s five-year, multi-institutional AIDE-QC project concentrates its efforts, will make quantum computers more functional and accessible to a broader group of researchers. The goal of AIDE-QC is to deliver a top-down open-source integrated software environment for quantum computers developed by a collaboration of scientists from national labs and academia that could democratize the power of this budding technology among researchers. After all, the number of scientists with highly complex scientific problems to simulate is much higher than the number of people who can operate a quantum computer.  

AIDE-QC is an amalgamation of several research “thrusts” that come together to improve the usability and functionality of quantum computing, de Jong notes. This includes programming languages that provide tools and a layer of abstraction that allow for more facile use of quantum computers, and compilers to communicate this code to the hardware. Other thrusts are more focused on smoothing out operations – namely verification and debugging, optimization, and a software integration thrust that helps tie these various threads together. 

According to de Jong, the work of AIDE-QC has already enabled quantum computers to do bigger, more efficient calculations with fewer qubits. “AIDE-QC’s highest-level goal is to enable any researcher who wants to run a simulation that requires a quantum computer to be able to do that without needing to know the ins and outs of quantum computing,” he explained.

AIDE-QC and a previous related project, QAT4Chem, focus on the noisy intermediate-scale quantum (NISQ) computing space, where they have already made significant contributions, for example for optimization by developing the SciKit-Quant library. NISQ, as the name implies, typically produces noisy data, in part because quantum calculations are probabilistic in nature, so effective optimization produces significant returns in the form of more accurate calculations on complex environments.

The AIDE-QC team has also notched achievements in the computation of free energies, extending the timescale of dynamic simulations, and scaling quantum synthesis through the development of its BQSKit Quantum Compiler and Synthesis toolkit. The group’s compilers are quite efficient and provide an important capability to the quantum computing environment. BQSKit and the SciKit-Quant library are being integrated into IBM’s Qiskit quantum computing code, while other commercial entities have shown interest in using or collaborating on the development of the software. High-energy physicists, nuclear physicists, and materials scientists are already starting to make good use of the software stack to run calculations that tend to be too cumbersome for conventional hardware, de Jong noted.

Simulating Natural Systems

Major advances in physics and chemistry that could revolutionize energy production, de-carbonization, drug development, cleaner industrial processes, and much more lie just on the other side of the quantum computing capabilities de Jong and his colleagues are advancing toward.

One such application that has been touted in de Jong’s field of computational chemistry is the process of converting nitrogen to ammonia for fertilizer. Right now, a significant portion of global energy consumption – something like three percent – is devoted to this process. But an enzyme cofactor called FeMoco performs this same function naturally, which could point the way toward a much more energy-efficient method.  

“We want to be able to simulate that [natural] system with all its dynamical behavior and understand why it works and how it works,” de Jong said. “But that’s a problem you cannot solve with a classical computer – it’s too big.”

There are many other applications in chemical and materials sciences waiting for scientific discovery with the help of quantum computers, he added, “such as increasing solar energy capture efficiency, developing longer-lasting batteries, and capturing carbon to reduce greenhouse gases.”

The AIDE-QC project brings the scientific community closer to achieving one of the core promises of quantum computing: to understand natural phenomena, particularly quantum phenomena, in a way that is both more holistic and granular.

As Richard Feynman noted, “it is impossible to represent the results of quantum mechanics with a classical universal device.”

When systems become too large, complex, and dynamic for classical computers to handle, the standard method is to segment or approximate systems to keep them simple enough for our current hardware.

“A quantum computer gives us the potential to study bigger problems that are more realistic to nature,” de Jong said.