Li Jiancong, Tian Jing, Lin Yudeng, Zhou Zhiwei, Li Yi, Gao Bin, Tang Jianshi, Chen Jia, He Yuhui, Qian He, Wu Huaqiang, Miao Xiangshui
School of Integrated Circuits, Hubei Key Laboratory for Advanced Memories, Huazhong University of Science and Technology, Wuhan 430074, China.
School of Integrated Circuits, Beijing National Research Center for Information Science and Technology (BNRist), Tsinghua University, Beijing, China.
Sci Adv. 2025 Jun 20;11(25):eadv4446. doi: 10.1126/sciadv.adv4446.
Emerging artificial intelligence for science (AI-for-Science) algorithms, such as the Fourier neural operator (FNO), enabled fast and efficient scientific simulation. However, extensive data transfers and intensive high-precision computing are necessary for network training, which challenges conventional digital computing platforms. Here, we demonstrated the potential of a heterogeneous computing-in-memristor (CIM) system to accelerate the FNO for scientific modeling tasks. Our system contains eight four-kilobit memristor chips with embedded floating-point computing workflows and a heterogeneous training scheme, representing a heterogeneous CIM platform that leverages precision-limited analog devices to accelerate floating-point neural network training. We demonstrate the capabilities of this system by solving the one-dimensional Burgers' equation and modeling the three-dimensional thermal conduction phenomenon. An expected nearly 116 times to 21 times increase in computational energy efficiency was achieved, with solution precision comparable to those of digital processors. Our results extend in-memristor computing applicability beyond edge neural networks and facilitate construction of future AI-for-Science computing platforms.
新兴的科学人工智能（AI-for-Science）算法，如傅里叶神经算子（FNO），实现了快速高效的科学模拟。然而，网络训练需要大量的数据传输和密集的高精度计算，这对传统数字计算平台构成了挑战。在此，我们展示了异构忆阻器计算（CIM）系统在加速FNO进行科学建模任务方面的潜力。我们的系统包含八个具有嵌入式浮点计算工作流程和异构训练方案的4千位忆阻器芯片，代表了一个利用精度有限的模拟设备来加速浮点神经网络训练的异构CIM平台。我们通过求解一维伯格斯方程和对三维热传导现象进行建模来展示该系统的能力。实现了计算能量效率预期近116倍至21倍的提升，其求解精度与数字处理器相当。我们的结果将忆阻器计算的适用性扩展到边缘神经网络之外，并促进了未来科学人工智能计算平台的构建。
