The prototypes of quantum computers currently available are gradually breaking down some barriers. These machines have a weak point: they make mistakes. This is the reason why Ignacio Cirac, the Spanish physicist who, together with Peter Zoller, developed the theoretical basis of quantum computing, holds that the correct thing is to identify them as prototypes to differentiate them from the fully functional quantum computers that will hopefully arrive in the future.
During the conversation we had with Ignacio Cirac in June 2021, the director of the Theoretical Division of the Max Planck Institute for Quantum Optics he explained to us who believed that quantum computers will be very valuable tools in the field of quantum chemistry to, for example, design drugs. Just five years after that conversation, a very important milestone has occurred that invites us to scan the horizon of this discipline with a very healthy optimism.
And a group of researchers from IBM; the RIKEN Center for Quantum Computing, in Japan; and Cleveland Clinic, in the USA, have carried out the largest quantum-classical chemistry simulation carried out to date. It’s a very important achievement for a reason: it represents a huge leap in the way quantum computers can be used alongside classical supercomputers to study real-world chemistry problems.
“This result is a dream”
Dr. Kenneth Merz, the leader of this research, assures that the result obtained by the team he leads is a dream. Until now, the most ambitious simulation that had been possible in this area using a quantum computer recreated a protein with only 303 atoms. However, Merz’s team has managed to simulate two biologically relevant proteins (T4-Lysozyme and Trypsin), as well as the molecules to which they bind, in a completely realistic aqueous environment and reaching 12,635 atoms.
To make this possible, they have used two quantum processors that add up to 94 qubits, executing 9,200 circuits over more than 100 hours and collecting 1.3 billion measurement results. The quantum data were subsequently processed with the Japanese supercomputer Fugaku. In this area, the calculation capacity of quantum computers makes a difference, although the merit does not belong exclusively to these machines.
The strategy that these scientists have developed consists of dividing large molecules into smaller, more manageable groups.
The strategy these scientists have developed is to divide large molecules into smaller, more manageable groups. Classical supercomputers solve the simpler regions, while quantum systems address the more complex and computationally demanding parts.
The results are then recombined to obtain a global image of the molecule. To carry out this simulation, the researchers introduced improvements in both classical and quantum techniques. However, one of the most important innovations they have developed is the improvement of the way in which the system identifies which parts of a molecule require detailed quantum treatment, which reduces the overall computational cost.
As we have just seen, we are facing a very important milestone, although we need to put it in context. And, despite its value, the strategy that these researchers have developed still does not surpass the best classical approaches. However, it demonstrates that quantum systems can already contribute to the resolution of significant scientific problems, especially when integrated with existing computing infrastructure.
Image | IBM
More information | Interesting Engineering
In Xataka | Beyond AI, US semiconductor manufacturers face the real battle of the future: quantum chips
GIPHY App Key not set. Please check settings