抛开电子，这项突破利用光-物质粒子驱动人工智能

* *Key terms:* ENIAC (埃尼阿克/ENIAC), Penn researchers (宾夕法尼亚大学研究人员), J. Presper Eckert, John Mauchly (names need transliteration or standard usage), launch the modern computing era (开启了现代计算时代), streams of electrons (电子流), powers (驱动/提供动力), demanding (需求日益增长/要求越来越高), electron-based hardware (基于电子的硬件). * *Draft:* 由宾夕法尼亚大学研究人员 J. Presper Eckert 和 John Mauchly 开发的 ENIAC，利用电子流解决复杂的数学问题，从而开启了现代计算时代。这种同样的电子学方法至今仍驱动着当今的计算机、智能手机和人工智能系统。但随着人工智能的需求日益增长，基于电子的硬件局限性正变得愈发难以忽视。 * **Para 2:** "Why Electrons Are Reaching Their Limits" * *Draft:* **电子为何正触及极限** * **Para 3:** "Electrons carry an electrical charge, which creates several challenges inside modern computer chips. As they move through materials, they generate heat and face resistance that wastes energy. Those problems become even more difficult as chips grow more complex and process enormous amounts of data for AI applications." * *Key terms:* electrical charge (电荷), computer chips (计算机芯片), resistance (电阻), wastes energy (能量损耗). * *Draft:* 电子携带电荷，这在现代计算机芯片内部引发了一些挑战。当它们穿过材料时，会产生热量并面临电阻，从而造成能量浪费。随着芯片变得越来越复杂，并为人工智能应用处理海量数据，这些问题变得更加棘手。 * **Para 4:** "Researchers led by Penn physicist Bo Zhen in the School of Arts & Sciences believe photons, the particles that make up light, could help solve some of these issues." * *Key terms:* Penn physicist Bo Zhen (宾夕法尼亚大学物理学家甄博 - need to check standard translation for Bo Zhen, usually 甄博 or Zhen Bo. Given "Penn", it's likely University of Pennsylvania. "School of Arts & Sciences" -> 文理学院), photons (光子). * *Draft:* 宾夕法尼亚大学文理学院物理学家甄博领导的研究人员认为，构成光的粒子——光子，或许有助于解决其中一些问题。 * **Para 5:** "'Because they are charge-neutral and have zero rest mass, photons can carry information quickly over long distances with minimal loss, dominating communications technology,' explains Li He, co-first author of a paper published in Physical Review Letters and a former postdoctoral researcher in the Zhen Lab. 'But that neutrality means they barely interact with their environment, making them bad at the sort of signal-switching logic that computers depend on.'" * *Key terms:* charge-neutral (电中性), zero rest mass (静止质量为零), minimal loss (极小的损耗), dominating (主导), Physical Review Letters (《物理评论快报》), co-first author (共同第一作者), postdoctoral researcher (博士后研究员), signal-switching logic (信号开关逻辑/信号切换逻辑). * *Draft:* “由于光子呈电中性且静止质量为零，它们能以极小的损耗长距离快速传输信息，因此在通信技术领域占据主导地位，”发表在《物理评论快报》上的一篇论文的共同第一作者、Zhen 实验室前博士后研究员何立解释道。“但这种中性也意味着它们几乎不与环境发生相互作用，导致它们难以胜任计算机所依赖的信号开关逻辑。” * **Para 6:** "In other words, light is excellent for carrying information quickly and efficiently, but it struggles with the switching operations needed for computing." * *Draft:* 换言之，光非常擅长快速高效地传输信息，但在执行计算所需的开关操作方面却显得力不从心。 * **Para 7:** "Combining Light and Matter for AI Computing" * *Draft:* **结合光与物质用于人工智能计算** * **Para 8:** "To overcome that problem, Zhen's team developed a special quasiparticle called an exciton-polariton. The particle forms when photons are strongly linked with electrons inside an atomically thin semiconductor material. This combination allows light to interact much more effectively, making it capable of performing the signal switching required for computing tasks." * *Key terms:* quasiparticle (准粒子), exciton-polariton (激子-极化激元), atomically thin semiconductor material (原子级薄半导体材料). * *Draft:* 为了克服这一问题，甄博的团队开发了一种被称为激子-极化激元的特殊准粒子。这种粒子是在光子与原子级薄半导体材料内的电子发生强耦合时形成的。这种结合使光能够更有效地进行相互作用，从而使其具备执行计算任务所需的信号开关能力。 * **Para 9:** "The breakthrough could be especially important for artificial intelligence systems, which consume enormous amounts of power." * *Draft:* 这一突破对于耗电量巨大的人工智能系统可能尤为重要。 * **Para 10:** "Many experimental photonic AI chips already use light to handle certain calculations at high speed. However, when these systems need to perform nonlinear activation steps, such as decision-making operations, they usually must convert light signals back into electronic ones. That conversion slows the process and increases energy use, reducing the benefits of photonic computing." * *Key terms:* photonic AI chips (光子人工智能芯片), nonlinear activation steps (非线性激活步骤), decision-making operations (决策操作), convert (转换). * *Draft:* 许多实验性的光子人工智能芯片已经利用光来高速处理某些计算。然而，当这些系统需要执行非线性激活步骤（如决策操作）时，通常必须将光信号转换回电信号。这种转换会减慢处理速度并增加能耗，从而削弱了光子计算的优势。 * **Para 11:** "Using exciton-polaritons, the Penn researchers demonstrated all-light switching while using only about 4 quadrillionths of a joule of energy. That amount is extraordinarily small, far below the energy needed to briefly power a tiny LED light." * *Key terms:* all-light switching (全光开关), 4 quadrillionths of a joule (4千万亿分之一焦耳 - quadrillion is 10^15, so 4 x 10^-15 joules, or 4 飞焦 femtojoules. "4 quadrillionths of a joule" translates literally to 4千万亿分之一焦耳). * *Draft:* 利用激子-极化激元，宾夕法尼亚大学的研究人员展示了全光开关，且仅消耗约 4 千万亿分之一焦耳的能量。这一能量极小，远低于短暂点亮微小 LED 灯所需的能量。 * **Para 12:** "Toward Faster and More Efficient AI Chips" * *Draft:* **迈向更快、更高效的人工智能芯片** 3. **Review and Refine:** * *Names:* J. Presper Eckert (J·普雷斯珀·埃克特), John Mauchly (约翰·莫奇利). These are famous historical figures, better use standard Chinese names. * *Bo Zhen:* 甄博. * *Li He:* 何立. * *Unit check:* "4 quadrillionths of a joule". Quadrillion in short scale (US/Modern British) is $10^{15}$. So $4 \times 10^{-15}$ J. In Chinese, $10^{-15}$ is 飞 or 千万亿分之一. "4千万亿分之一焦耳" is accurate and readable. * *Flow:* The text flows well. The distinction between "electronic" (电子的) and "photonic" (光子的) is clear. "Exciton-polariton" is standardly translated as "激子-极化激元". "Quasiparticle" is "准粒子". 4. **Final Polish (Self-Correction during drafting):** * Para 1: "J. Presper Eckert and John Mauchly" -> J·普雷斯珀·埃克特和约翰·莫奇利. * Para 4: "School of Arts & Sciences" -> 文理学院. * Para 8: "strongly linked" -> 强耦合 is the technical term for exciton-polariton formation. "Strongly linked" is a bit layman, but "强耦合" is the precise scientific term implied. I will use "强耦合" or "紧密结合" but "强耦合" fits the physics context better. Let's stick to 如果该技术能成功实现规模化，它将有望催生出能够直接处理相机信息、无需在光与电之间反复转换的光子芯片。这种方法还可能降低大型人工智能系统巨大的能源需求，并有望在未来芯片上支持基本的量子计算功能。 甄波（Bo Zhen）是宾夕法尼亚大学文理学院物理与天文系Jin K. Lee总统副教授。 何力（Li He）曾任宾夕法尼亚大学文理学院甄波实验室的博士后研究员。他现任蒙大拿州立大学助理教授。 该研究的其他作者包括来自宾夕法尼亚大学文理学院的王志（Zhi Wang）和Bumho Kim。 这项研究得到了美国海军研究办公室（N00014-20-1-2325 和 N00014-21-1-2703）以及斯隆基金会的资助。