A biocomputer that would use 10,000 times less energy than a conventional computer

⇧ [VIDÉO] You may also like this partner content (by ad)

The problem with the use of traditional electronic processors for information processing remains the excessive consumption of electricity. Therefore, turning to biology to solve computer problems seems to be an interesting solution to save energy. Researchers have just designed a biocomputer the size of a chip that uses molecules that move in channels to (quickly) solve complex problems.

According to some researchers, the fast-growing but still energy-intensive information and communication technology sector is expected to consume 20% of global electricity production by 2030. This does not mean that the energy efficiency of electronic computers will not improve, but slow down. by heat production, small-scale (eg quantum) effects and increased costs.

For ten years, biocomputers have emerged as an interesting alternative for reducing energy consumption for computers. As a reminder, biological computers use molecules of biological origin – such as DNA and proteins – to perform numerical or real calculations. In particular, the development of nanobiotechnology (on a nanometric scale) has enabled researchers to design specific biomolecular systems, which may ultimately lead to the computational functionality of the computer.

Using this alternative method, Till Korten from Dresden University of Technology (Germany) and his colleagues have built a chip-based biocomputer that uses molecules that travel in channels to solve problems. ” Network-based bioinformatics is a very energy-efficient approach that makes it possible to encode a combinatorial problem in a modular graphics network embedded in a nanofabricated device. “, The researchers report. Since a combinatorial problem has a large number of variables, electronic computers would take billions of years to solve.

Kinesins operates microtubules with high energy efficiency

The physical network – the computer chip – is then explored using the biology of a fluid that contains fibers that make up the cell’s cytoskeleton (microtubules) and kinesins. It is a motor protein that makes it possible to propel molecules, such as microtubules, forward with great energy efficiency. The researchers explain that the advantage of microtubules is that they all move at the same time, making it possible to perform calculations simultaneously, in addition to being very accessible. The microtubules move through the channels, and each path they take is equivalent to the computer trying to solve a given problem.

© Surendiran, Meinecke, Salhotra et al. (2022)

As a result, we have estimated that network-based bioinformatics uses several orders of magnitude less energy per second. operation than an electronic computer. “, the researchers write. In fact, each step in the calculation would require 10,000 times less energy to move the molecules in a biocomputer than the electrons in a conventional computer. In addition, compared to other biocomputers using the same technology, the network-based one would perform 128 times fewer calculations.

However, in order to demonstrate the viability of this new technology, it will be necessary to demonstrate applicability to other issues of practical importance and significant upscaling of the technology. The researchers also want to use more molecules to speed up the solution of the problem, although it could lead to more errors in the microtubules.

Source: ACS Nanoscience

Leave a Comment