Scientists have developed a new energy efficient biological supercomputer that could one day replace the world's most powerful machines.
Supercomputers have the power to carry out unimaginably complex computations, but there is one major problem – they use a ton of energy. According to a report from Business Standard, however, scientists have made a stunning new development that could solve the energy problem for supercomputers once and for all.
Researchers have made a model biological supercomputer, which could process information quickly and accurately using parallel networks like a traditional supercomputer would. It runs on biological processes similar to the ones that power all living cells.
According to Dan Nicolau from McGill University in Montreal, “We’ve managed to create a very complex network in a very small area.” The scientists took a number of cues from the way nature transfers energy, using Adenosine triphosphate, or ATP, as the main source powering the new biological supercomputer.
Researchers combined geometrical modeling and the latest in computer engineering to create a largely capable machine on a seriously tiny scale. The computer hasn’t replaced some of the world’s top supercomputers just yet, but the demonstration that it actually works is a huge step forward.
The main chip in the biological supercomputer resembles a birds eye view of a busy city, and measures only 1.5 cm across. The major development was switching out traditional microchips that move electrons with a charge for a new chip that uses short protein strings called biological agents which travel around in the circuit, propelled by ATP.
The benefit of using biological agents is that there is little heat generated as the computer runs. The new chips allow supercomputers to operate using significantly less energy than before. We may be leaving the days where a supercomputer needs its own power plant to function behind.
The super computer was proficient at solving a complex classical mathematical problem using parallel computing in the same way a normal supercomputer would, but researchers say they still have a long way to go before it can take on Watson and the like.
A press release from McGill University describing the details of the study can be found here.