DNA computer reaches answers through thermodynamic balance

Irish researchers have developed a new type of molecular computer that solves complex mathematical problems through thermodynamic equilibrium, ERR reports. The DNA-based system moves towards the correct answer on its own.

Modern electronics and engineered biological systems consume substantial amounts of energy when performing calculations. To prevent errors, researchers have traditionally relied on complex and resource-intensive kinetic control mechanisms. The new computer was designed so that the correct answer corresponds to the system’s most stable state with the lowest energy.

Computer energy consumption has exceeded the amount theoretically required for calculations for decades. Earlier molecular computers effectively had to work against the current because energy had to be continuously pumped into the machines to prevent errors. According to research group leader Damien Woods, the gap between computers’ actual and theoretical energy consumption provides a strong incentive to rethink the design of existing technology.

The new DNA computer resembles a self-assembling puzzle. Its backbone is a long, one-dimensional scaffold to which specialised computational blocks attach in sequence. These tiny components, which function like a programming language, use molecular interactions to find the physically best-fitting positions for themselves.

Blocks that attach in the wrong place form weak connections and soon fall out of the system. This means the computer does not need a separate error-correction system. The final answer emerges only when all the components occupy the correct positions and their connections fully match. Co-author Tristan Stérin said the approach produced a conceptually simple and highly error-resistant information technology tool.

Adding and multiplying

The researchers wanted to determine the computer’s performance limits, so they gave it complex mathematical calculations to solve. To build a network several times larger than earlier ones, they borrowed material directly from nature. The core of the new scaffold was a long DNA strand extracted from the harmless M13 bacteriophage.

Using this biological foundation, the device accurately added 25-bit numbers. When processing longer numbers, the molecules had to work through a 100-bit sequence of calculations. The computer also detected data parity, multiplied values by three and divided ternary units by two.

According to co-author Constantine Glen Evans, the model’s scalability demonstrates its real capability. Although the computer was given difficult mathematical problems, it reached stable final answers at the microscopic level remarkably quickly. For simpler algorithms, the molecules settled into a stable final answer in less than 60 seconds.