过去的几十年，各大公司都在芯片上暗暗较劲：芯片涨价、GPU短缺、AI算力焦虑... 就在大家盯着芯片迭代升级时，微软在悄悄做另一件事：用光重新定义计算。 他们花了四年，用手机摄像头、Micro LED和透镜，拼出了一台模拟光学计算机（AOC）。 如今，这个实验已经登上Nature，带来了一个足以颠覆GPU的未来想象。 光子登场：固定点搜索的秘密 几十年来，算力的故事几乎都写在硅片上：摩尔定律的加速、GPU的堆叠、能耗的焦虑。 可在英国剑桥，微软研究院的一支小团队走了一条完全不同的路——让光来算数。 他们拼出了一台模拟光学计算机（AOC），材料一点也不稀有：Micro LED、光学镜头、还有来自手机的摄像头传感器。 看上去更像是一台实验室「组装机」，却打开了算力的另一种可能。 英国剑桥Microsoft Research实验室模拟光学计算机的详细图像。它是使用市售部件制造的，例如micro-LED灯和智能手机摄像头的传感器 其实，光学计算的设想早在20世纪60年代就被提出过，只是在当时受限于工艺，一直停留在理论层面。 如今，微软团队把它真正做了出来。 AOC真正的秘密不在这些零件，在于它的运行方式——固定点搜索 ...

Core Insights - A groundbreaking study from the University of Utah's College of Engineering introduces a method to encode partial differential equations (PDEs) into light waves, processed by a novel optical device called the Optical Neural Engine (ONE), marking a significant step from theoretical exploration to practical application in optical computing [1][2]. Group 1: Technology and Methodology - The ONE system combines the advantages of diffractive optical neural networks and optical matrix multipliers, modeling PDEs using optical methods rather than traditional digital representations [1]. - The system utilizes different properties of light waves, such as intensity and phase, to represent various variables in the equations, allowing the light signals to evolve through a series of optical components to yield solutions to specific PDEs [1][2]. Group 2: Applications and Impact - The research tested the ONE system on several classical PDEs, including Darcy's law (for groundwater flow modeling), the static magnetic Poisson equation in demagnetization processes, and the Navier-Stokes equations (widely used in fluid mechanics), demonstrating good adaptability and accuracy [2]. - This research provides a multifunctional, high-efficiency platform for large-scale scientific computing and engineering simulations, with potential significant impacts in fields such as geological modeling, chip design, and climate simulation [2].

Core Viewpoint - China is leading in foundational research for next-generation computing, raising concerns that U.S. export controls may become ineffective if these research results are commercialized [1][4]. Group 1: Research Output - From 2018 to 2023, China published 160,852 semiconductor-related papers, more than double the U.S. output of 71,688 papers [1]. - China's semiconductor research papers grew by 41% during this period, significantly outpacing India (26%), the U.S. (17%), and South Korea (6%) [1]. - In terms of impactful research, China authored 23,520 papers in the top 10% of citations, nearly half of the total, compared to the U.S. with 10,300 papers [2]. Group 2: Institutional Strength - Among the top 10 institutions publishing semiconductor research from 2018 to 2023, 9 are Chinese [2]. - The report from the Korea Institute of Science and Technology Evaluation and Planning indicates that South Korea's semiconductor technology capabilities are rated lower than China's across various fields [3]. Group 3: Future Implications - U.S. export controls on advanced semiconductors and manufacturing equipment may struggle to contain China's growth as it shifts focus to new semiconductor architectures [4]. - New fields such as neuromorphic computing and optical computing are identified as key growth areas for China's semiconductor research [4]. - Analysts predict that if China successfully commercializes next-generation semiconductor technologies, it could not only catch up to but potentially surpass the U.S. [5].