In computer language and, making a very basic summary, the system works in a binary way, like our neurons. While the latter close or open to let electrical impulses, in computer sciences the binary systemIt is based on zeros and some: the zeros do not miss the current and the ones, yes. But … what would happen if there were more options?
In an advance that could redefine the future of computer science, a team of Chinese scientists has developed the first carbon -based microchip of the world capable of executing tasks of Artificial intelligence using a novel ternary logic systemovercoming current technology dependent on binary. Achievement marks a fundamental step towards next -generation semiconductors that could overcome silicon in speed, efficiency and scalability.
The team, led by Peng Lianmao, presented A carbon nanotubes (CNT) transistor chip that processes data not only in some and zeros, but also in a third statewhich allows calculations to be performed faster and less energy. In experiments, the chip demonstrated enormous precision in image recognition tasks. The results have been published in Science Advances.
The CNT are tiny tubes made of graphene rolled sheets. They have exceptional mechanical and electrical properties, as well as excellent stability, and are mainly used as conductive additives in lithium -ion batteries.
However, due precisely to their electrical characteristics and their ultra -ethro structure, CNT is also a promising material for next -generation semiconductor technology. In Comparison with traditional chips based In silicon, the CNT offers a higher speed, lower energy consumptionYoOh the possibility of TamañOS from transistors toornmtoS smallñyou: key advantages that you canYoto help overYoMITES OF THE MOORE LAW, which predicts a duplicationeithern of the ability of semiconductors every two years.
Most modern electronic devices are based on binary logical systems, where all data is represented by zeros and some. Therefore, the introduction of a third state could Allow more efficient data processing, which results in faster calculations without increasing circuit complexity. This approach, known as ternary logic, improves the efficiency of information transmission within the same physical space.
The research team designed a new CNT transistor, introducing a concept known as transistors with a source gate (SGT). By adjusting the gate voltage, CNT transistor can move between three different current statesforming the basis for ternary logical circuits.
On this basis, the Lianmao team built a neuronal network capable of learning and reasoning imitating the connections between human brain neurons. To check the capacities of the ternary logical circuit, they carried out thorough experiments, demonstrating that this CNT -based neuronal network achieved 100% accuracy In the classification of handwritten digits, which highlights its immense potential in artificial intelligence applications.
This technology has broad applications in high performance computing, artificial intelligence, automatic learning and low consumption storage devices. Also It could be used on IoT devices, where energy efficiency is a key concern.
CNT chips are still behind silicon chips in terms of integration density. For example, the NVIDIA RTX 5090 GPU, announced in January 2025, contains the amazing amount of 92 billion transistors, far beyond the existing capabilities of CNT technology.
However, the race for the based chips in carbon is A new border for the world semiconductor industry and could help save Moore’s law that is already reaching its limit.
“Our ultimate goal is to make The chips based in carbon generalize in the PReitherXimos 10 to 15 Añyou, What to allowto Its generalized use in supercomputers, data centers, Telandsmart phones and other electronic deviceseitherconventional only, ”concludes Peng.