算力商业
aibusiness.com ·
@Esther Shittu
收购Hugging Face,Nvidia在AI竞赛中取得重大胜利
评论认为,尽管关注焦点在模型制造商,但Nvidia通过收购Hugging Face成为当前AI竞赛的明星。
AI推荐理由 评论Nvidia收购Hugging Face在AI竞赛中的意义,认为Nvidia是当前明星。分析性内容,提供产业视角。
70
正文
生成式人工智能
基础模型
新闻
通过收购Hugging Face,英伟达在人工智能竞赛中取得重大胜利
虽然大多数注意力都集中在Anthropic、OpenAI和谷歌等模型制造商身上,关注哪家供应商将赢得竞赛,但此刻的明星是英伟达。
埃斯特·希图,新闻撰稿人,AI Business
2026年9月3日
4分钟阅读
Hugging Face
英伟达决定以近130亿美元收购开源人工智能平台Hugging Face,这标志着人工智能市场的一个关键时刻,并可能使这家AI硬件和软件巨头成为AI竞赛的领跑者。
这家供应商周四透露,计划扩展Hugging Face的平台,加强其基础设施,并扩大全球开发者和机构对AI的访问。英伟达表示,Hugging Face将继续支持开源和开放权重模型,以及多云和多加速器的开发和部署。
“我们有动力和动机保持这种开放和中立,访问最好的模型并连接最好的模型,”英伟达企业AI副总裁贾斯汀·博伊塔诺在关于收购的媒体简报会上说。
此次收购正值开源AI竞争加剧之际,因为中国出现了更多开源模型。随着阿里巴巴、月之暗面和DeepSeek发布开放模型,这些模型被认为在性能上与Anthropic、OpenAI和谷歌的前沿模型相当,美国开源市场出现了空白。英伟达一直在尝试通过其Nemotron开源模型系列慢慢填补这一空白。
相关:OpenAI称GPT-6 Astra是其最安全的模型,但它仍然危险
“这是一个强烈的信号,表明开放模型正在成为一个主要的商业市场,而不仅仅是开发者运动,”RPA2AI Research的首席执行官兼创始人卡夏普·科姆佩拉说。“英伟达实际上是在整个模型生态系统中进行对冲。”
他补充说,英伟达似乎押注,如果Anthropic、OpenAI或其他专有提供商在AI模型竞赛中领先,英伟达最终将在其目前凭借GPU主导的计算市场中获胜。另一方面,如果开放模型制造商占上风,英伟达将处于更有利的地位,因为它将拥有Hugging Face。
“这可能会增加模型竞争而不是减少竞争,特别是当企业采用更小、更专业的开放模型,而不是在每个用例中都依赖前沿模型时,”科姆佩拉说。
Hugging Face的中立性面临风险
然而,尽管此次收购凸显了开源在AI市场中的重要性,但像英伟达这样的巨头拥有如此庞大的开源平台可能会引发紧张。Hugging Face一直是一个平台,任何开发者都可以在上面发布新模型。中国供应商通常在该平台上发布产品。据Hugging Face联合创始人兼首席执行官克莱姆·德兰格称,其目标是在平台上达到1亿构建者并保持开放。截至8月,Hugging Face表示其存储库中有近300万个公共模型。
相关:提示:英伟达在AI堆栈中向上移动
“几乎从定义上讲,我们所做的一切都是开放的,”德兰格在媒体活动中说。他补充说,他希望该平台将继续“对生态系统中的所有参与者产生影响,让他们做任何想做的事情,并将这项技术分发给全世界每个人。”
科姆佩拉表示,Hugging Face保持中立的能力将取决于信任。他说,AMD、英特尔、谷歌、亚马逊和其他基础设施提供商等供应商现在可能会以不同的方式看待Hugging Face。
“中立性现在必须被证明,而不是被假定,”科姆佩拉说。“所有权不会自动消除中立性,但它确实转移了举证责任。以前来自Hugging Face作为独立中介地位的东西,将越来越多地必须来自可证明的行为。”
他补充说,GitHub,另一个大型开源平台,可以作为英伟达如何保持中立的例子。微软自2018年以来一直拥有该平台,但它仍然可信。
对中国的反响
此次收购对中国也有重大影响,DeepSeek和阿里巴巴等供应商定期在Hugging Face上发布内容。科姆佩拉表示,中国一直在积极尝试减少对英伟达的依赖,现在Hugging Face被英伟达拥有,中国供应商将有更多动力建立自己的模型存储库和基础设施硬件。
相关:企业AI初创公司Wonderful现在估值50亿美元
“他们不太可能立即放弃Hugging Face,因为其全球开发者覆盖范围仍然有价值,但对其的战略依赖将更难合理化,”他说。
对企业意味着什么
英伟达与Hugging Face的结合也可能增加企业的集中风险,这些企业可能最终过度依赖英伟达,包括其GPU、CUDA GPU计算架构、网络、推理软件,以及现在的Hugging Face。
“英伟达可能使开放AI更容易采用,同时使供应商锁定成为更大的担忧,”科姆佩拉说。
然而,最终,此次收购可能使英伟达成为AI竞赛中最强大的竞争者,因为它现在不仅拥有一个模型,而是拥有一个模型平台。
“拥有平台的公司不需要构建获胜的应用程序,也能从获胜的应用程序中受益,”科姆佩拉说。“英伟达正越来越多地在AI领域以同样的方式定位自己。”
关于作者
埃斯特·希图
新闻撰稿人,AI Business
埃斯特·希图自2021年以来一直报道AI技术和行业趋势。作为Targeting AI播客的联合主持人,她与专家、思想领袖和实践者探讨关键的AI发展。在加入AI Business之前,她曾为SearchEnterpriseAI、纽约每日新闻、Bklyner和布鲁克林每日鹰报撰稿。当她不深入AI世界时,她会花时间在激情项目上并抚养她的三个女儿。
查看更多埃斯特·希图的文章
想要在谷歌搜索结果中看到更多AI Business故事?
立即添加我们
订阅新闻通讯
将最新的AI新闻和见解直接发送到您的收件箱!
保持更新!
查看原文(中文正文为 AI 翻译)
Generative AI
Foundation Models
News
With Hugging Face Acquisition, Nvidia Scores Big Win in AI Race
While most attention has been on model makers such as Anthropic, OpenAI and Google regarding which vendor will win the race, the star of the moment is Nvidia.
Esther Shittu,News Writer, AI Business
September 3, 2026
4 Min Read
Hugging Face
Nvidia’s decision to acquire open source AI platform Hugging Face for nearly $13 billion marks a pivotal moment in the AI market and could make the AI hardware and software giant the frontrunner in the AI race.
The vendor on Thursday revealed that it plans to scale Hugging Face’s platform, strengthen its infrastructure and expand access to AI for developers and institutions worldwide. Nvidia said that Hugging Face will continue to support open source and open-weight models and multi-cloud and multi-accelerator development and deployment.
“We are incentivized and motivated to keep this open and neutral, accessing the best models and connecting the best models,” said Justin Boitano, vice president of enterprise AI at Nvidia, during a media briefing about the acquisition.
The acquisition comes at a time when competition in open source AI has revved up following the emergence of more open source models coming out of China. With Alibaba, Moonshot AI and DeepSeek releasing open models that are considered comparable in performance with frontier models from Anthropic, OpenAI and Google, a gap has developed in the open source market in the U.S. Nvidia had been slowly trying to fill that space with its Nemotron open source model line.
Related:OpenAI Touts GPT-6 Astra as Its Safest Model, But It's Still Dangerous
“This is a strong signal that open models are becoming a major commercial market, not just a developer movement,” said Kashyap Kompella, CEO and founder of RPA2AI Research. “Nvidia is effectively hedging across the model ecosystem.”
He added that Nvidia appears to be betting that if Anthropic, OpenAI or another proprietary provider pulls ahead in the AI model race, Nvidia will end up winning in the compute market it currently dominates with its GPUs. On the other hand, if open model makers prevail, Nvidia will be in an even stronger position because it will own Hugging Face.
“This could increase model competition rather than reduce it, particularly as enterprises adopt smaller and more specialized open models instead of relying on frontier models for every use case,” Kompella said.
Hugging Face’s Neutrality at Stake
However, while the acquisition accentuates the importance of open source in the AI market, a giant like Nvidia owning such a large open source platform could create tension. Hugging Face has been a platform on which any developer can release new models. Chinese vendors usually release their products on the platform. And according to Clem Delange, co-founder and CEO of Hugging Face, its goal is to reach 100 million builders on the platform and keep it open. As of August, Hugging Face said it had nearly three million public models in its repository.
Related:Prompt: Nvidia Moves up the AI Stack
“Almost by definition, everything we do is open,” Delange said during the media event. He added that he hopes the platform will continue to be “impactful for all the players in the ecosystem to do whatever they want to and distribute this technology for everyone all over the world.”
Hugging Face's ability to remain neutral will depend on trust, Kompella said. He said that vendors such as AMD, Intel, Google, Amazon and other infrastructure providers could view Hugging Face differently now.
“Neutrality will now have to be demonstrated rather than assumed,” Kompella said. “Ownership does not automatically eliminate neutrality, but it does shift the burden of proof. What previously came from Hugging Face’s position as an independent intermediary will increasingly have to come from demonstrable behaviour.”
He added that GitHub, another big open source platform, could serve as an example of how Nvidia could maintain neutrality. Microsoft has owned the platform since 2018, yet it remains credible.
Repercussions for China
The acquisition also has significant implications for China, with vendors such as DeepSeek and Alibaba publishing regularly on Hugging Face. China has been actively trying to reduce its dependence on Nvidia, and with Hugging Face now owned by Nvidia, Chinese vendors will have even more incentive to build their own model repositories and infrastructure hardware, Kompella said.
Related:Enterprise AI Startup Wonderful Now Valued at $5B
“They are unlikely to abandon Hugging Face immediately because its global developer reach remains valuable, but strategic dependence on it will become harder to justify,” he said.
What it Means for Enterprises
The Nvidia-Hugging Face combination could also increase concentration risk for enterprises, which could end up overly dependent on Nvidia, with its GPUs, CUDA GPU computing architecture, networking, inference software, and now, Hugging Face.
“Nvidia may make open AI easier to adopt while simultaneously making vendor lock-in a bigger concern,” Kompella said.
Ultimately, though, the acquisition could make Nvidia the strongest competitor in the AI race because it now doesn't own just one model but a platform of models.
“The company that owns the platform does not need to build the winning application to benefit from the applications that win,” Kompella said. “Nvidia is increasingly positioning itself in the same way across AI.”
About the Author
Esther Shittu
News Writer, AI Business
Esther Shittu has covered AI technologies and industry trends since 2021. As co-host of the Targeting AI podcast, she talks with experts, thought leaders and practitioners exploring critical AI developments. Before AI Business, she wrote for SearchEnterpriseAI, the New York Daily News, Bklyner and the Brooklyn Daily Eagle. When she's not diving deep into the world of AI, she spends her time on passion projects and raising her three daughters.
See more from Esther Shittu
Want more AI Business stories in your Google search results?
Add Us Now
Sign Up for the Newsletter
The most up-to-date AI news and insights delivered right to your inbox!
Stay Updated!