Ask Indrakshi Raychowdhury what a quantum computer is bad at, and she points to arithmetic.
“If you try to add two numbers, a classical calculator does it easily, whereas doing it on a quantum computer is incredibly difficult,” she said in a conversation with The Hindu. “But simulating a highly complex, interacting quantum system (where particles mutually influence each other) is relatively easy for a quantum computer. That is highly counterintuitive, but it is where quantum advantage lies.”
“Understanding exactly where quantum computers can provide this advantage is currently the biggest challenge, and it is why governments and companies are investing heavily in this space,” Dr. Raychowdhury, an associate professor in the physics department at BITS Pilani (Goa campus), added.
High bar
The problem her team solved a few weeks ago is of the useful kind. According to one expert assessment, it is only one of three problems that are incredibly useful to physicists and also impervious (so far) to the muscle of classical computers. In other words, it demonstrates ‘quantum advantage’.
Working with scientists from IBM Quantum, Dr. Raychowdhury and her colleagues simulated the behaviour of subatomic particles on 120 qubits — qubits being the computing currency of quantum computers — of an IBM processor. The calculation took the quantum machine 20 seconds and a classical computer two hours.
The Quantum Advantage Tracker (QAT), the expert assessor, exists to police that claim. Launched in November 2025 by IBM, with the Flatiron Institute, the startups BlueQubit and Algorithmiq, and a widening cast of academic reviewers, QAT is a kind of live peer review where a researcher submits a result and a committee reproduces it — on both classical and quantum machines — before letting it stand. The bar is high. Of some 220 items logged on its GitHub page, only around a dozen have been accepted, and just a handful survive as ‘active’ claims.
The Quantum Advantage Tracker homepage.
| Photo Credit:
Quantum Advantage Tracker
“It is more thorough than a standard academic journal,” Dr. Raychowdhury said. It is here that her group’s entry now sits, marked ‘active’ — the first such from an Indian laboratory. Were the problem to be solvable by a classical machine, it would be considered ‘superseded’. How long their solution stays ‘active’ is an open challenge: since May, they haven’t been superseded.
Looking for the right problem
For all the money and noise around quantum computing, the machines have yet to solve a genuinely useful problem faster than an ordinary computer. The best-known demonstrations have been contrived. When Google reported in 2019 that its processor had finished in minutes a task a supercomputer would need 10,000 years to match, the task had no practical use and was chosen precisely because it suited quantum hardware. (To be sure, the QAT has a space for those too.)
Breaking modern encryption or simulating a large molecule from scratch remains years away. So the field has begun to chase a narrower prize: “useful” advantage, a quantum win on a real problem whose answer can be independently checked.
Dr. Raychowdhury’s problem of choice lies at the foundations of physics. The quarks inside a proton, and the gluons that bind them, are governed by the strong nuclear force, and working out how they move and rearrange in real time is among the toughest problems in the field. Classical methods are good at snapshots — they can pin down a proton’s mass with great accuracy — but watching the particles evolve, frame by frame, is “exponentially hard to calculate classically,” she said. A quantum computer, being itself made of quantum parts, can instead be coaxed into imitating the system directly: a still photograph replaced by a film.
The quarks inside a proton and the gluons that bind them are governed by nature’s strongest force.
| Photo Credit:
Image created with AI
“I am actually visiting CERN (Switzerland) next month,” she said, where the world’s most powerful particle smasher pulverises protons to physically trace how the simplest particles may evolve into weightier, more complicated ones. CERN researchers, hemmed in by the limits of classical methods, have invited her, she said, to know whether her approach can carry their simulations further.
‘Without algorithms, hardware is useless’
The model her team used was a simplified stand-in for the full theory of the strong force, which is nature’s most powerful force. While no new particle has emerged from it, the novelty lies in the method. Her group’s encoding stripped away mathematical redundancy so the simulation could scale; a task others had run on at most 27 qubits was pushed to 120. “If someone gave me a 1,000-qubit computer today, our algorithm would scale to it seamlessly,” she reckoned.
Just as portable, she believes, is the trick her team used to beat the noise that plagues fragile quantum hardware: run the experiment twice — once with the particle, once without — within fractions of a second, then subtract one measurement from the other, so that the near-identical errors cancel out. “It is an inexpensive way of mitigating noise,” she said, and colleagues have already begun applying it to other problems. None of it required her to leave her desk. IBM’s machines are reached over the cloud, programmed in Python through a toolkit called Qiskit, and most of her runs finished in minutes. “While the media often highlights quantum applications in drug discovery or material science, those fields are at a much more primitive stage computationally than high-energy physics.”
India is only beginning to build towards this. Its National Quantum Mission, the ₹6,003-crore programme approved in 2023, is funding quantum algorithms and runs a computing hub in Bengaluru, but the country has no high-end quantum hardware of its own, and researchers here lean on machines abroad.
“Without algorithms, the hardware is useless; you can’t just sit in front of a quantum computer and get an answer,” she said.
jacob.koshy@thehindu.co.in
Published – July 09, 2026 09:00 am IST
