A Chinese research team has developed an innovative computing architecture that nearly quadruples computing power, opening up new possibilities in areas such as embodied intelligence , edge sensing , brain-inspired computing and communication systems.The Fourier Transform acts as a frequency “translator,” converting complex signals, including sound and images, into frequency domain representations, a fundamental method widely used in science and engineering.
Targeting this universal computing process, a research team from Peking University creatively integrated two new devices suitable for frequency conversion into a multi-physics-domain architecture , resulting in a versatile hardware system capable of performing a wide range of operations, including the Fourier Transform.
Their breakthrough was published in the journal Nature Electronics, Friday (9/1).
"This architecture allows different computing paradigms to operate in optimal physical domains, such as electric current, charge, or light, thereby improving computational efficiency," said Tao Yaoyu, a researcher at the university's Institute for Artificial Intelligence (AI).
Tao said the integrated system leverages the complementary strengths of both devices in frequency generation, modulation, and in-memory computing. While maintaining accuracy and reducing power consumption, the system increases the Fourier Transform computation speed from about 130 billion operations per second to about 500 billion, a several-fold increase.
Such computing architectures can help enable new hardware to operate efficiently and accelerate its adoption in a wide range of areas, including basic AI models, embodied intelligence, autonomous driving, brain-computer interfaces, and communication systems.
