How China Is Girding for an AI Battle With the U.S.
中国如何备战与美国的 AI 竞赛

Many of the initiatives were on display at an AI conference that ended this week in Shanghai, which Chinese authorities used as a showcase for products free of U.S. technologies. 
本周在上海闭幕的一场人工智能大会上，多项自主创新成果集中亮相。中国官方将这场展会作为展示摆脱美国技术产品的平台。

One startup, Shanghai-based StepFun, touted a new AI model that it said required less computing power and memory than other systems, making it more compatible with Chinese-made semiconductors. Although Chinese chips are less capable than American products, Huawei Technologies and other companies have been narrowing the gap by clustering more chips together, boosting their performance.
总部位于上海的初创企业阶跃星辰(StepFun)展示了一款新型 AI 模型，宣称其所需算力和内存低于其他系统，因而更适配国产芯片。虽然中国芯片性能仍逊于美国产品，但华为等企业正通过多芯片集群技术提升性能，逐步缩小差距。

China also released an AI global governance plan at the event, the World Artificial Intelligence Conference, which called for establishing an international open-source community through which AI models can be freely deployed and improved by users. Industry participants say it showed China’s ambition to set global standards for AI and could undermine the U.S., whose leading models aren’t open-source.
在世界人工智能大会期间，中国还发布了全球人工智能治理倡议，呼吁建立国际开源社区，使用户能自由部署和改进 AI 模型。业内人士指出，这彰显中国欲主导 AI 国际标准制定的雄心，可能对拒绝开源核心模型的美国形成制衡。

The conference followed a series of announcements and investments in China aimed at turbocharging its AI capabilities, including rapid expansions in power generation and skills training.
这场会议之前，中国已宣布了一系列旨在加速人工智能能力建设的举措和投资，包括快速扩充发电能力和技能培训规模。

The whole-nation effort, led by Beijing, includes billions of dollars in spending by state-owned enterprises, private companies and local governments. 
在中央政府主导下，这场举国行动涉及国有企业、私营企业和地方政府高达数十亿美元的支出。

China’s securities regulator has largely refrained from approving initial public offerings of companies whose businesses aren’t related to “hard tech” sectors such as semiconductors and AI, so that capital can concentrate on funding technologies of strategic importance, people involved in the process said.
知情人士称，中国证券监管部门基本暂停批准与半导体和人工智能等"硬科技"领域无关的企业进行首次公开募股，以便资本能集中投向具有战略意义的技术领域。

Victory in the global AI race carries high stakes. AI is expected to upend economies and militaries, and leadership in the sector is seen as critical to future global influence and national security.
全球人工智能竞赛的胜利事关重大。人工智能预计将颠覆经济和军事领域，而该领域的领导地位被视为对未来全球影响力和国家安全至关重要。

The U.S. maintains its early lead, with Silicon Valley home to the most popular AI models and the most powerful chips. Much of China’s AI spending has led to waste and overcapacity. 
美国保持着早期的领先优势，硅谷拥有最受欢迎的人工智能模型和最强大的芯片。中国在人工智能领域的大量投入导致了资源浪费和产能过剩。

China also clearly wants U.S. technologies. Access to advanced chips has been a priority for Chinese negotiators in trade talks, The Wall Street Journal has reported. Sales of Nvidia’s H20 AI chips to China were recently restored by Washington after it restricted them in April, a reversal seen by Beijing as a gesture of good faith in the talks.
中国显然也渴望获得美国技术。据《华尔街日报》报道，在贸易谈判中获取先进芯片一直是中国谈判代表的首要任务。美国政府四月份对英伟达 H20 人工智能芯片实施出口限制后，近期又恢复了对华销售，北京方面将这一转变视为美方在谈判中展现善意的姿态。

The Trump administration is taking steps to try to preserve the U.S.’s lead. It recently announced an AI “action plan,” which aims to slash red tape and make it easier for tech companies to build data centers needed to train AI models.
特朗普政府正采取措施试图保持美国的领先地位。近日宣布了一项人工智能"行动计划"，旨在减少繁文缛节，使科技公司更容易建立训练 AI 模型所需的数据中心。

Earlier this year at the White House, OpenAI and Japan’s SoftBank unveiled a $500 billion effort to build new AI data centers, though the project has faced delays.
今年早些时候，OpenAI 和日本软银在白宫宣布了一项耗资 5000 亿美元建设新型人工智能数据中心的计划，不过该项目已遭遇延期。

China, though, has shown a willingness to spend whatever it takes. The rising popularity of DeepSeek, the Chinese AI startup, has buoyed Beijing’s hopes that it can become more self-sufficient. Huawei has published several papers this year detailing how its researchers used its homegrown chips to build large language models without relying on American technology.
然而，中国已展现出不惜代价投入的决心。中国 AI 初创企业深度求索的迅速走红，增强了北京对实现技术自主的信心。华为今年发布的多篇论文详细阐述其研究人员如何利用国产芯片构建大语言模型，完全摆脱对美国技术的依赖。

“China is obviously making progress in hardening its AI and computing ecosystem,” said Michael Frank, founder of think tank Seldon Strategies.
智库塞尔登战略创始人迈克尔·弗兰克表示："中国显然正在强化其人工智能和计算生态系统方面取得进展。"

‘Project Spare Tire’  “备胎计划”

China’s biggest AI challenge is overcoming its difficulty in sourcing the world’s most-advanced chips. Washington has cut China off from some of Nvidia’s most sophisticated semiconductors, as well as the advanced machine tools used to make cutting-edge chips, restrictions that many experts believe will continue to hold China back.
中国在人工智能领域面临的最大挑战是难以获取全球最先进的芯片。美国政府已切断中国获取英伟达部分最尖端半导体及用于制造前沿芯片的先进设备的渠道，许多专家认为这些限制将持续阻碍中国的发展。

Huawei is helping spearhead efforts to navigate those restrictions. During a meeting with President Xi Jinping in February, Chief Executive Officer Ren Zhengfei told Xi about “Project Spare Tire,” an effort by Huawei and 2,000 other enterprises to help China’s semiconductor sector achieve a self-sufficiency rate of 70% by 2028, according to people familiar with the meeting. 
华为正带头努力突破这些限制。据知情人士透露，在今年 2 月与习近平主席的会晤中，华为首席执行官任正非向习主席汇报了"备胎计划"——华为联合 2000 家企业共同推进的项目，旨在帮助中国半导体产业到 2028 年实现 70%的自给率。

Increasingly, the company has been able to bundle together the best chips it can produce to match the performance of some American computing systems. That is helping local companies reach some of the same computing goals as the U.S., such as training state-of-the-art generative AI models, though it consumes more power than U.S. chips. 
该公司已能够将自产的最佳芯片组合起来，以匹配某些美国计算系统的性能。这帮助本土企业实现了与美国类似的计算目标，例如训练最先进的生成式人工智能模型，尽管其能耗高于美国芯片。

U.S. researcher SemiAnalysis recently reported that one such Huawei cluster, which connects 384 of its Ascend chips, outperforms Nvidia’s flagship system with 72 graphics-processing units on some metrics. 
美国研究机构 SemiAnalysis 近期报告显示，由 384 颗昇腾芯片组成的华为集群在部分指标上超越了搭载 72 块图形处理单元的英伟达旗舰系统。

Morgan Stanley analysts forecast that China will have 82% of AI chips from domestic makers by 2027, up from 34% in 2024.
摩根士丹利分析师预测，到 2027 年中国国产 AI 芯片占比将从 2024 年的 34%提升至 82%。

China’s government has played an important role, funding new chip initiatives and other projects. In July, the local government in Shenzhen, where Huawei is based, said it was raising around $700 million to invest in strengthening an “independent and controllable” semiconductor supply chain.
中国政府发挥了重要作用，为芯片新项目和其他计划提供资金支持。华为总部所在地深圳的当地政府 7 月表示，正筹集约 7 亿美元资金用于投资强化"自主可控"的半导体供应链。

Prodded by Beijing, Chinese financial institutions, state-owned companies and government agencies have rushed to deploy Chinese-made AI models, including DeepSeek and Alibaba’s Qwen. That has fueled demand for homegrown AI technologies and fostered domestic supply chains.
在中央政府的推动下，中国金融机构、国有企业和政府部门争相部署国产 AI 模型，包括深度求索的 DeepSeek 和阿里巴巴的通义千问。这既刺激了对本土 AI 技术的需求，也培育了国内供应链。

Some companies ordered expensive AI servers to deploy the models even before they came up with a use for the technology, according to people familiar with the matter. 
据知情人士透露，部分企业在尚未明确技术应用场景前，就已订购昂贵的 AI 服务器来部署这些模型。

Open-source advantage  开源优势

At the Shanghai conference, organized by China’s government and featuring more than 800 companies, Chinese researchers played down the impact of U.S. export controls on advanced chips. They compared notes on how they were focusing on overcoming bottlenecks through better model designs and engineering techniques.
在上海这场由政府主办、800 余家企业参与的会议上，中国研究人员淡化了美国先进芯片出口管制的影响。他们交流了如何通过优化模型设计和工程技术来突破瓶颈的经验。

Many also touted Chinese companies’ tendency to give users free access to modify and deploy their AI models, an open-source approach that has boosted adoption of Chinese models globally.  
许多人也称赞中国企业倾向于让用户免费修改和部署其 AI 模型，这种开源方式促进了中国模型在全球范围内的采用。

While the world’s best large language model is still American, the best model that everyone can use free is now Chinese. According to benchmark provider Artificial Analysis, the overall performance of China’s best open-weight model has surpassed the American champion since November.
尽管全球最强大的大语言模型仍属美国，但如今人人都能免费使用的最佳模型已来自中国。基准测试机构 Artificial Analysis 数据显示，自去年 11 月起，中国最佳开源权重模型的综合性能已超越美国冠军。

In recent weeks, a flurry of Chinese companies have flooded the market with open-source models, many of which are claiming to surpass DeepSeek’s performance in certain use cases. OpenAI’s Sam Altman said his company had pushed back the release of its open-source model indefinitely for further safety testing.
近几周来，大批中国企业纷纷向市场推出开源模型，其中许多宣称在特定应用场景下性能已超越深度求索。OpenAI 的萨姆·奥尔特曼表示，该公司已无限期推迟开源模型的发布以进行进一步安全测试。

Talent and power  人才与实力

China is also investing heavily in other areas, including more electricity to power domestic data centers to develop and run AI.
中国还在其他领域大力投资，包括为开发和运行人工智能的国内数据中心提供更多电力。

It is spending $564 billion on grid construction projects in the five years up to 2030, an increase of more than 40% from the previous five years, Morgan Stanley researchers forecast.  
摩根士丹利研究人员预测，中国将在截至 2030 年的五年内投入 5640 亿美元用于电网建设项目，较前五年增长逾 40%。

China currently has about 2.5 times as much power-generation capacity as the U.S., a disparity that is projected to grow larger in the next five years despite the U.S. expanding power generation.
中国目前的发电装机容量约为美国的 2.5 倍，尽管美国正在扩大发电规模，预计未来五年这一差距还将进一步扩大。

China has also approved more than 600 colleges to set up degree programs in AI, according to data released by the country’s Ministry of Education in April. That is up from 35 universities with such programs in 2019. 
根据中国教育部 4 月公布的数据，全国已有 600 多所高校获批设立人工智能专业学位点，较 2019 年开设该专业的 35 所高校实现大幅增长。

In Beijing, primary and secondary schools will begin mandatory AI lessons for students starting in September. 
北京市中小学将从今年 9 月起开设人工智能必修课程。

China’s efforts have already enabled it to develop a robust pipeline of homegrown talent, according to researchers from the Hoover Institution and Stanford University, who recently evaluated the backgrounds of more than 200 authors involved in DeepSeek’s papers between 2024 and February 2025. More than half of these DeepSeek researchers never left China for schooling or work, they found.
胡佛研究所与斯坦福大学研究人员近期评估了 2024 年至 2025 年 2 月期间参与深度求索公司论文撰写的 200 余位作者的背景后发现，中国培养的本土人才已形成稳定输送渠道。研究显示，超过半数深度求索的研究人员从未出国求学或工作。

The U.S. has fewer universities that offer AI degree programs, but American universities dominate rankings for computer and information science. This April, President Trump signed an executive order mandating AI education and learning opportunities for American youth.
美国开设人工智能学位课程的大学较少，但美国高校在计算机与信息科学领域排名占据主导地位。今年四月，特朗普总统签署行政命令，要求为美国青少年提供人工智能教育及学习机会。

Write to Raffaele Huang at raffaele.huang@wsj.com and Liza Lin at liza.lin@wsj.com
请发送邮件至 raffaele.huang@wsj.com 联系黄瑞福，或发送邮件至 liza.lin@wsj.com 联系李丽莎

The Global AI Race  全球人工智能竞赛

Coverage of advancements in artificial intelligence, selected by the editors
人工智能前沿进展精选报道（编辑严选）

Get WSJ's AI Newsletter  订阅《华尔街日报》AI 通讯

AI Boom Sparks Fight Over Soaring Power Costs
人工智能热潮引发电力成本飙升之争

The High-Schoolers Who Just Beat the World’s Smartest AI Models
击败全球最强大 AI 模型的高中生们

Trump Administration Pledges to Stimulate AI Use and Exports
特朗普政府承诺推动人工智能应用与出口

SoftBank and OpenAI Struggle to Launch $500 Billion Stargate AI Project
软银与 OpenAI 启动 5000 亿美元星际之门 AI 项目遇阻