Core Viewpoint - The research team from Peking University has developed a novel multi-physical domain fusion computing architecture that enhances the performance of Fourier transform calculations by nearly four times, utilizing new post-Moore devices [1] Group 1: Technological Innovation - The new architecture integrates volatile vanadium oxide devices with non-volatile tantalum/hafnium oxide devices, creating a hardware system suitable for diverse computational methods including Fourier transform [1] - The system integration allows for complementary advantages in frequency generation control and in-memory computing, increasing the speed of Fourier transform calculations from approximately 130 billion operations per second to about 500 billion operations per second [1] Group 2: Application Potential - This new computing framework is expected to address the challenges of expanding the operator spectrum of post-Moore devices, enabling support for multiple computational methods and accelerating the practical applications of these new devices in advanced fields such as artificial intelligence foundational models, embodied intelligence, autonomous driving, brain-machine interfaces, and communication systems [1]

Core Viewpoint - The research team from Peking University has developed a novel multi-physical domain fusion computing architecture that utilizes post-Moore new devices to support Fourier transforms, achieving nearly a fourfold increase in computing power, which opens new possibilities in fields such as embodied intelligence, edge perception, brain-like computing, and communication systems [1][2]. Group 1: New Computing Architecture - The new computing architecture allows various computing methods to operate in their suitable physical domains such as current, charge, and light, enhancing computational efficiency [2]. - The integration of volatile vanadium oxide devices and non-volatile tantalum/hafnium oxide devices leverages their complementary advantages in frequency generation control and storage-computation integration, improving the speed of Fourier transform calculations from approximately 130 billion operations per second to about 500 billion operations per second [2]. Group 2: Applications and Implications - The new framework is expected to overcome the challenges of expanding the operator spectrum of post-Moore new devices, enabling support for multiple computing methods, thus allowing new devices to function effectively [2]. - This advancement is anticipated to accelerate the application of new devices in cutting-edge fields such as artificial intelligence foundational models, embodied intelligence, autonomous driving, brain-machine interfaces, and communication systems [2].

Core Viewpoint - A new multi-physical domain fusion computing architecture developed by a research team from Peking University significantly enhances the performance of Fourier transforms, achieving nearly a fourfold increase in computing power, which opens new possibilities in various fields such as embodied intelligence, edge perception, brain-like computing, and communication systems [1][2]. Group 1 - The new computing architecture allows for multiple computing methods to be executed in their suitable physical domains, such as current, charge, and light, thereby improving computational efficiency [2]. - The integration of volatile vanadium oxide devices and non-volatile tantalum/hafnium oxide devices enables complementary advantages in frequency generation control and storage-computation integration, increasing the speed of Fourier transform calculations from approximately 130 billion operations per second to about 500 billion operations per second [2]. - This new framework is expected to overcome the challenges of expanding the operator spectrum of post-Moore new devices, allowing them to support various computing methods and accelerating their application in cutting-edge fields such as artificial intelligence foundational models, embodied intelligence, autonomous driving, brain-machine interfaces, and communication systems [2].

Core Insights - The research team from Peking University has developed a novel multi-physical domain fusion computing architecture that utilizes post-Moore new devices to support Fourier transforms, achieving nearly a fourfold increase in computing power, which opens new possibilities in fields such as embodied intelligence, edge perception, brain-like computing, and communication systems [1][2]. Group 1: Technological Advancements - The new computing architecture allows for various computational methods to be executed in their suitable physical domains, such as current, charge, and light, enhancing computational efficiency [2]. - The integration of volatile vanadium oxide devices and non-volatile tantalum/hafnium oxide devices has created a hardware system capable of diverse computational methods, including Fourier transforms [1]. Group 2: Performance Improvements - The new framework has improved the speed of Fourier transform calculations from approximately 130 billion operations per second to about 500 billion operations per second, representing a significant increase in computational speed [2]. - The innovative computing framework is expected to overcome the challenges of expanding the operator spectrum of post-Moore new devices, enabling them to support multiple computational methods effectively [2].