Professor Hossam Amroush, a researcher at the Technical University of Munich (TUM), has developed a new artificial intelligence technology that relies on storing data in memory. This technology is a significant improvement over current methods of using information technology in memory.
The new computing style relies on the use of active ferroelectric field switches (FeFETs), a different approach than previous chips. Instead of being used only for calculations, FEROK domain keys are now also used to store data. This means improving chip performance while saving time and energy.
This technology can be used in many fields, such as productive artificial intelligence, deep learning algorithms, and robotics applications. Professor Hossam Amroush believes that he will be able to achieve this goal within a few years.
The new chip’s concept is inspired by the architecture of chips in the human brain, where neurons process signals and store information at the same time. The chip uses ferroelectric switches that can store information even without a power source.
Although there are potential uses for this technology in areas such as running deep learning algorithms and processing aircraft data, it is expected to take a few years before the chips are ready for general use. Although performance and efficiency have improved, there are a number of challenges to overcome such as industry safety requirements.
FAQs:
What new AI technology has been developed?
The new technology for artificial intelligence relies on storing data in memory using efficient ferroelectric field switches (FeFETs).
What are the uses of this technology?
This technology can be used in areas such as productive artificial intelligence, deep learning algorithms, and robotics applications.
What is the main advantage of this technology?
The chips’ performance is improved while saving time and energy due to data being stored in memory.