From Molecules to the Cosmos — The Real Promise of Quantum Computing
Part 4 of “Quantum, explained simply” — but slightly deeper insight for decision-makers who want the real picture, not the hype. By Amit Agarwal, CEO & Co-Founder, SeQure AG.
This series has dwelt, of necessity, on the threat: Shor's algorithm breaking RSA, Grover nibbling at AES. So it is only fair to turn the coin over. The very same physics that endangers today's cryptography is also one of the most promising tools that humanity has ever built — for medicine, for materials, and, at its furthest edge, for the deepest scientific or even metaphysical questions we know how to ask. Powerful tools, as I noted in the companion post, do not pick sides. This article is the optimistic half of the story, told with the same discipline as the rest: what is genuinely near, what is merely promising, and what is gloriously, honestly speculative.
Why quantum helps at all — simulating nature
Recall the founding insight of the whole field, from the first article: a quantum computer is not a faster classical computer; it is a machine that speaks nature's own language. That matters most for a single, enormous task — simulating quantum systems themselves. The difficulty is brutal and precise: the information needed to describe a collection of interacting quantum particles grows exponentially with their number, so a classical computer drowns almost immediately. A quantum computer represents such a system natively.
|
The mathematics Describing n interacting quantum particles requires on the order of 2n numbers (the dimension of the combined state space). Classical cost therefore scales as classical ∝ 2n vs quantum ∝ n (native) Around fifty particles, 2n already exceeds the memory of the world's largest supercomputers. A quantum machine sidesteps the wall entirely — which is exactly the use Feynman proposed in 1982. Figure 1. The exponential wall that makes quantum chemistry hard classically — and the reason quantum computers exist. |
The machine is becoming real
For decades the obstacle was not the idea but the hardware: quantum states are fragile, and noise corrupts them faster than computation can proceed. The answer is quantum error correction — spreading one reliable "logical" qubit across many imperfect physical ones. Until recently this remained a promise. In 2024 it became a demonstrated fact. Google's "Willow" processor showed, for the first time convincingly, that adding more physical qubits to a larger error-correcting code makes the logical qubit better, not worse — the hallmark of operating below the so-called error-correction threshold.
|
The mathematics Below the threshold, the logical error rate falls geometrically as the code distance d grows. In Google's result each step suppressed the error by a factor Λ ≈ 2.14 per increase in code distance. A suppression factor greater than 1 is the whole game: it means scale now helps. It does not mean a useful machine exists yet — but it removes the deepest doubt about whether one can. Figure 2. The 2024 milestone: errors fall as the system grows. A necessary foundation for everything that follows. |
Honesty requires a caveat here too. The headline "quantum advantage" demonstrations so far — sampling tasks a quantum chip performs faster than a supercomputer — are on artificial benchmarks chosen to be hard, not on problems anyone needs solved. The achievement is real; its present commercial value is essentially nil. What has changed is the trajectory: error correction works, and the roadmap to machines of a million physical qubits is now an engineering programme rather than a hope.
Where the value lands first: chemistry, materials, medicine
Because simulation is quantum computing's natural strength, the most credible near-term value lies in chemistry and materials science. Designing a molecule, a catalyst or a battery electrode means understanding quantum behaviour that classical methods can only approximate. Early, concrete steps are already visible: AstraZeneca, with AWS, IonQ and NVIDIA, has demonstrated a quantum-accelerated computational-chemistry workflow for reactions used in drug synthesis; researchers at Phasecraft have devised an algorithm reported to need on the order of a million-fold fewer steps to model a promising battery-electrode material. None of this is yet outperforming the best classical methods in production — the honest word remains "promising" — but the industries that live or die by molecular discovery are paying close attention. One widely cited estimate, from McKinsey, puts the potential value to life sciences and chemicals in the hundreds of billions of dollars by 2035; treat such figures as direction, not prophecy.
The promising, unproven middle: optimisation and finance
One step further from certainty sits a large family of optimisation problems — routing, scheduling, and, closer to home, portfolio construction, hedging and risk simulation. The hope is that quantum or quantum-inspired methods will find better solutions, or find them faster, than classical optimisation. The reality, stated plainly, is that a clear, durable quantum advantage for these tasks has not yet been demonstrated in a commercially viable manner; classical algorithms are a fast-moving target. This is an area where a lot of companies and research institutions are investing time and money in especially using the power of quantum machine learning and far more enhanced Monte-Carlo simulations. The initial results in controlled test environments are promising but still not ready to be commercialised-at-scale.
The infrastructure leaves the planet
A more unexpected frontier is physical rather than algorithmic — and it was a recurring theme at Quantum Industry Day in Geneva in 2025, which I had the privilege to attend. Quantum hardware craves cold, stillness and isolation, and some of the most demanding of those conditions exist naturally in space. The first photonic quantum processor reached orbit in June 2025 — a University of Vienna experiment roughly the size of a shoebox, operating some 550 kilometres up. In parallel, a wave of ventures is building orbital data centres that rent computing capacity from space: Axiom Space has flown a data-centre prototype to the International Space Station, and NVIDIA-backed efforts plan orbital clouds of accelerators. The logical next step — already being discussed seriously — is to host quantum processors on such platforms, letting institutions rent quantum time on hardware that floats in the cold and quiet of orbit. It is early, and space imposes its own penalties (radiation chief among them), but the idea has moved from the conference corridor to first hardware with striking speed.
An honest map of where this stands
It helps to keep the layers separate, because conflating them is how hype is manufactured. Nearer to us, and more certain, is simulation. Further out, and less proven, is optimisation. And at the horizon — profound, thrilling and frankly speculative — sit the deepest questions in physics.
|
THE MATHEMATICS Figure 3. A map worth carrying into a strategy discussion: the higher the tier, the sooner and surer; the lower, the more profound and more speculative. |
The deepest frontier: physics, the cosmos, and the nature of reality
Here the wonder is real, and the caution must be equal to it. Quantum computers are already finding use as laboratory instruments for fundamental physics: simulating the "lattice gauge theories" that describe the forces inside atomic nuclei, and even modelling aspects of black-hole physics through so-called holographic lattices, in which the entanglement structure of a quantum system mirrors the geometry of spacetime itself. These are genuine research results — early, narrow, but real — and they hint at using quantum machines to probe conditions we can never recreate in a laboratory, from the interior of a black hole to the first instants after the Big Bang.
Bolder thinkers go further still. A growing strand of theoretical work asks whether information is not merely something we compute with, but a fundamental constituent of reality itself. Florian Neukart of Terra Quantum, for example, has proposed a "Quantum Memory Matrix," in which spacetime acts as an information-bearing medium that records the events passing through it — a framework he connects to enduring mysteries such as dark matter and dark energy. And at the most speculative edge of all, the same information-first worldview leads some, like Neppe and Close with their “Triadic Dimensional Distinction Vortical Paradigm (TDVP)”, to ask whether computation and consciousness might be two faces of a single phenomenon. I want to be precise about the status of these ideas: they are frontier theories, with supporting mathematics, but needs more heavy lifting before they can be considered established science. But that is exactly what makes them worth knowing about. The history of physics is, in part, a history of questions that looked like philosophy until the mathematics caught up. If do not theorize, you do not get to the theory of relativity, or to quantum computing for that matter.
|
THE MATHEMATICS Figure 4. Illustrative and frontier: the same quantum tools, pointed from the smallest scale to the largest questions we know how to ask. |
The takeaway
So, is quantum computing a threat or a gift? The honest answer is that it is one technology with two faces, and a serious institution should keep both in view. The threat to cryptography is concrete, near enough to demand preparation, and the subject of the rest of this series. The promise — better medicines and materials first, then perhaps a genuine deepening of our understanding of the universe — is the reason the field exists at all, and the reason it will keep advancing whether or not any single bank is ready. The mature posture is neither dread nor dazzlement: prepare diligently for the risk, and stay curious about the upside. At SeQure AG we help financial institutions handle the first with rigour — precisely so that they are free to be excited, rather than blindsided, by the second. The quantum era is coming either way. It is a remarkable time to be paying attention.
——————————————————————————————
References & notes
1. R. P. Feynman, "Simulating Physics with Computers," International Journal of Theoretical Physics 21, 467–488 (1982).
2. Google Quantum AI, "Quantum error correction below the surface code threshold," Nature (2024) — the "Willow" result; logical-error suppression factor Λ ≈ 2.14 per code-distance step.
3. AstraZeneca with AWS, IonQ and NVIDIA — demonstration of a quantum-accelerated computational-chemistry workflow for drug-synthesis reactions (2025); company announcements.
4. Phasecraft — quantum algorithm for modelling battery-electrode materials (e.g. strontium vanadate), reported ~106-fold reduction in steps (2025); reported in the trade and scientific press.
5. McKinsey & Company, analysis of quantum computing in life sciences and chemicals (2025) — market estimate, cited as an industry projection, not a verified figure.
6. University of Vienna (P. Walther et al.) — first photonic quantum processor operated in low-Earth orbit (2025); and orbital-data-centre initiatives (Axiom Space; NVIDIA-backed ventures) — press and company sources.
7. Quantum simulation of lattice gauge theories, and "lattice holography" studies relating entanglement to spacetime geometry (Ryu–Takayanagi) — recent peer-reviewed physics literature (e.g. Physical Review D, 2024).
8. F. Neukart (Terra Quantum / Leiden University) — the "Quantum Memory Matrix" framework and related writings on information, dark matter/dark energy, and consciousness.
9. Neppe, V. M., & Close, E. R. (2020). The Neppe-Close triadic dimensional vortical paradigm: An invited summary. International Journal of Physics Research and Applications, 3(1), 001–014.
About the author. Amit Agarwal is CEO and Co-Founder of SeQure AG, a Swiss quantum cybersecurity company helping banks and financial institutions identify, prioritise, and remediate cryptographic vulnerabilities before Q-Day. He brings 25+ years across software, SaaS, payments and FinTech and holds a BTech (Computer Science), an MBA, an MAS, and a Quantum Computing qualification from MIT's executive education.