Since the early days of computing, humans have relentlessly pushed the boundaries of what machines can do. From Alan Turing’s theoretical computing machines to the silicon revolution of the late 20th century, we’ve watched our tools evolve from vacuum tubes to supercomputers capable of simulating the birth of stars. Yet even the fastest supercomputers today, from IBM’s Summit to Japan’s Fugaku, are subject to the fundamental constraints of classical physics. Their processors calculate one instruction after another—swiftly, yes, but still linearly—bounded by binary logic.
Quantum computers, however, threaten to flip this paradigm entirely on its head.
Instead of bits that represent either 0 or 1, quantum computers are built upon qubits—quantum bits that can represent 0 and 1 simultaneously, thanks to a quantum property known as superposition. They can also share a kind of spooky interdependence through entanglement, a phenomenon Albert Einstein famously referred to as “spooky action at a distance.” With these properties, quantum computers don’t just compute—they explore vast dimensions of possibility in parallel.
And that ability might change everything—from chemistry and material science to logistics, finance, and cryptography.
But there’s one big problem: proving that quantum computers are actually better at solving real-world problems than their classical counterparts—a concept known as quantum advantage—has been frustratingly elusive.
Until now.
Cracking the Quantum Code: The Legacy of Shor’s Algorithm
The theoretical promise of quantum advantage dates back three decades. In 1994, Peter Shor, a Caltech graduate working at Bell Labs, introduced an algorithm that could factor large numbers exponentially faster than the best-known classical methods.
Why does this matter?
Because factoring large numbers underpins much of modern cryptography—specifically, RSA encryption. The difficulty of breaking RSA encryption is what keeps our emails private, our bank accounts secure, and our digital identities safe. For classical computers, factoring a 2,048-bit number would take millions of years. Shor’s algorithm, theoretically, could break it in seconds using a quantum machine.
Shor’s insight sparked a quantum gold rush. Governments, research labs, and companies like IBM, Google, and Microsoft poured billions into developing practical quantum computers. And yet, nearly three decades later, we still haven’t built a machine capable of running Shor’s algorithm at the scale needed to break real-world encryption.
Worse still, researchers have struggled to find other problems where quantum computers could demonstrate a clear and compelling advantage.
Until recently, Shor’s algorithm remained the only shining beacon of quantum supremacy in the realm of practical computation.
That is, until a team led by Caltech stepped in with a new approach.
The Energy Landscape of the Universe—and How Quantum Computers Navigate It Better
In a landmark study published in Nature Physics, a Caltech-led team of researchers may have cracked open an entirely new frontier for quantum advantage—one that touches not just cryptography, but the deepest roots of physics, chemistry, and materials science.
The problem? Understanding how materials cool down to their most stable states, a challenge both deceptively simple and computationally profound.
“To a physicist,” says John Preskill, one of the pioneers of quantum information theory and a co-author of the paper, “cooling a material to its ground state is like watching nature solve an incredibly complex optimization problem.”
When you cool down a metal or a crystal or a gas, its atoms and electrons rearrange themselves into the lowest-energy configuration they can find. This final arrangement—called the ground state—determines the material’s properties: how it conducts electricity, how strong it is, how it interacts with light.
But here’s the twist: on the way to this ground state, the system might get temporarily stuck in a local minimum—a configuration that’s lower in energy than the surroundings, but not the absolute lowest possible. It’s like descending a mountain range and getting trapped in a valley that’s not the deepest point.
“For classical computers, finding the global minimum among a landscape of local minima is incredibly difficult,” says Preskill. “They get stuck.”
Quantum computers, by contrast, can use their inherent weirdness—superposition and tunneling—to leap past these plateaus. They can “sense” multiple valleys at once and find better paths down the energy landscape.
This is the essence of the breakthrough.
A New Quantum Algorithm, Tailored for Physical Reality
The Caltech team didn’t just stumble across this insight—they built a new quantum algorithm from the ground up to prove it.
Led by co-authors including Hsin-Yuan (Robert) Huang and Chi-Fang (Anthony) Chen, and backed by foundational insights from Fernando Brandão and John Preskill, the researchers developed a method that can identify these low-energy states more effectively than any known classical algorithm.
This isn’t just a theoretical footnote—it’s a practical framework for studying condensed matter systems, quantum chemistry, materials engineering, and high-energy physics.
“Physicists and chemists want to know how materials behave at their most stable configurations,” says Huang. “If we can compute the energy states of complex molecules or crystalline structures more accurately, we can predict everything from superconductivity to drug-binding behavior.”
And because the algorithm has been mathematically proven to outperform classical methods under specific conditions, it offers a new benchmark for quantum advantage.
Unlike Shor’s algorithm, which targets a very narrow problem (factoring), this new method opens doors to a broad class of simulations, from superconductors to particle interactions, making it one of the most promising use cases for quantum computing to date.
Experiment Meets Theory: Quantum Advantage in the Lab
In a second paper published alongside the theoretical study, another Caltech team brought this idea into the lab—proving that it’s not just possible to simulate local minima, but to physically create them in real materials.
David Hsieh and Gil Refael, working with a crystal of calcium ruthenate (Ca₂RuO₄), used an ultrashort burst of light—less than a trillionth of a second long—to shake the crystal out of its ground state and into a metastable local minimum.
In this new state, the electron spins—usually aligned in an antiparallel pattern—snapped into parallel alignment and stayed that way for microseconds, far longer than anyone expected. This demonstrated, in vivid detail, how local minima can be accessed, controlled, and potentially harnessed to transform a material’s properties on demand.
“These experiments bring to life what the quantum algorithms are trying to simulate,” says Preskill. “They’re two sides of the same coin.”
Why This Matters: From Materials to Medicine
If we can efficiently simulate how matter behaves at the quantum level, the implications are staggering.
Imagine designing materials with perfect conductivity for lossless power grids. Or creating next-generation batteries that charge in seconds and never degrade. Or accelerating drug discovery by simulating how thousands of potential molecules bind to disease targets—without ever touching a lab bench.
Today, these simulations are painstakingly slow, imprecise, and computationally expensive. Quantum algorithms like the one developed at Caltech could change that—rendering molecular and material simulations not just possible, but fast and reliable.
This is true quantum advantage, rooted not in abstract mathematics but in the core machinery of nature.
Quantum’s Next Leap: Building the Hardware
Of course, none of this matters if we can’t build the machines to run these algorithms. Quantum computers today are still in their infancy, plagued by noise, decoherence, and scalability issues. Most systems can only maintain a handful of qubits reliably, while real-world simulations may require thousands or even millions of qubits.
But progress is accelerating.
Companies like Google, IBM, Rigetti, IonQ, and startups around the globe are racing to overcome these challenges. Google’s Quantum AI team, for example, has already demonstrated quantum supremacy in narrow tasks like random circuit sampling. IBM has released a roadmap promising fault-tolerant machines within the next decade. Caltech’s own AWS Center for Quantum Computing, where Preskill and Brandão are leading research, is working on scalable quantum architectures rooted in superconducting qubits.
With each step forward in error correction, qubit stability, and architecture design, we inch closer to a machine that can run the algorithms theorized by Preskill’s team—and finally fulfill the promise of quantum computing.
Final Thoughts: A New Scientific Renaissance
We are standing at the threshold of a new computational era—one that will not only redefine how we solve problems but will fundamentally reshape our understanding of reality itself.
In the same way the microscope revealed the microbial world and the telescope unveiled the galaxies, quantum computers will unlock hidden layers of nature—energy landscapes, molecular dances, and cosmic equations that classical minds can’t even approximate.
The Caltech study isn’t just another paper in a prestigious journal—it’s a signal flare from the future. It tells us that quantum advantage is real, measurable, and perhaps most importantly, useful.
All that remains is to build the machines that can turn theory into reality.
The countdown to the quantum age has begun.
References: Chi-Fang Chen et al, Local minima in quantum systems, Nature Physics (2025). DOI: 10.1038/s41567-025-02781-4
Xinwei Li et al, Time-hidden magnetic order in a multi-orbital Mott insulator, Nature Physics (2025). DOI: 10.1038/s41567-024-02752-1