Quick Summary of the Key Points
On September 3, NVIDIA officially announced the acquisition of Hugging Face, the world's largest AI open-source community, for $12.93 billion. This transaction directly draws parallels to classic industry cases such as Microsoft's acquisition of the code hosting platform GitHub in 2018 and IBM's acquisition of the open-source software vendor Red Hat. It highlights a long-standing hidden rule in the industry: almost all open-source technology communities that have built their success on free sharing struggle to generate revenue on their own and ultimately either cease to exist or are absorbed by giants. This acquisition serves both parties' needs—Hugging Face faces challenges in commercialization, and by selling off, it can secure a stable income; NVIDIA, on the other hand, gains control over the global AI developer community, providing a sustainable growth boost for its graphics card business.
---
Detailed and Easy-to-Understand Explanation
1. This $12.9 billion acquisition is not about a mere website; it's about the “default starting point” for all AI practitioners
Many may not be familiar with Hugging Face, but essentially, it's like a free marketplace for the AI community. Whether individual developers want to create small AI applications or large companies want to develop their own models, their first stop is Hugging Face to find ready-made open-source models, optimized code, and learn from others' experiences. It has always been a recognized neutral platform where anyone can freely download and share resources without any barriers. NVIDIA's investment is not just in the servers and team; it's in the usage habits of all AI professionals. From now on, the first step in any AI development project will involve using Hugging Face, and this platform is now completely under NVIDIA's control.
2. Open-source communities, despite their widespread popularity, find it harder to make money than traditional businesses
Just because Hugging Face is well-known doesn't mean it generates substantial profits. According to reports, its annual revenue is only $150 million, which is just over 1 billion RMB, and it has not yet turned a profit. The fundamental flaw in the business model of open-source communities is that while they can attract millions of users with free content, they can only make money by offering value-added services (e.g., providing model optimization support to companies or charging membership fees). However, developers are known for preferring free options, so even a slight increase in subscription fees often leads to a mass switch to alternative platforms. GitHub, for example, never made a profit before being acquired by Microsoft, and the Chinese open-source community, OpenChina, nearly went bankrupt before being saved by Baidu. This shows that 99% of open-source communities cannot survive on their own. Hugging Face's sale for $12.9 billion is equivalent to more than 80 years of its total revenue, and its previous investors have made a substantial profit with no loss.
3. Giants acquiring open-source communities is not about charity; it's about gaining control of the industry's “free information network”
You might think it's a waste for a giant to spend billions on a non-profit platform, but such deals are much more profitable in the long run. Microsoft acquired GitHub to drive traffic to its cloud services, and IBM acquired Red Hat to build an ecosystem for its hybrid cloud business. NVIDIA's acquisition of Hugging Face is even more lucrative: all AI companies, whether they are NVIDIA's customers or competitors aiming to compete with NVIDIA's graphics cards, will have to upload their new models to Hugging Face for testing. This allows NVIDIA to monitor the industry's progress in real-time, understand the latest developments, and prepare or optimize its products accordingly. The value of this information is far beyond the cost of the acquisition.
4. Can open-source communities remain neutral after being acquired by giants? Unlikely; they will likely remain open on the surface but biased in practice
Many developers are concerned that Hugging Face might start favoring NVIDIA's hardware after the acquisition. However, past cases show that giants don't completely lose the neutrality of these platforms. For example, after Microsoft acquired GitHub, the industry continued to thrive. NVIDIA has also promised to maintain Hugging Face's openness and avoid exclusivity. However, it's almost inevitable that there will be some bias—for instance, models uploaded to Hugging Face might be optimized for NVIDIA GPUs, making other platforms less competitive. This type of gradual influence is much more subtle than direct monopolization.
5. The open-source scene in China has already followed this path: almost all major platforms are backed by large companies
This acquisition further reflects the current state of the open-source community in China. Well-known platforms like Alibaba’s MagPa and Baidu’s PaddlePaddle are all supported by large companies. The only independent open-source community, OpenChina, nearly went bankrupt before being rescued by Baidu, which acquired a 60% stake in the company. It struggled to become profitable on its own and only recently achieved stability. In China, developers are less willing to pay for open-source tools compared to their overseas counterparts. Without the support of large companies, independent open-source communities rarely survive more than three years. It's likely that the future of China's AI open-source scene will be dominated by a few major platforms, with little room for neutral third-party platforms.
---
This translation maintains the structure and tone of financial journalism, adapting the content to fit the target audience's cultural and linguistic context while ensuring the accuracy and consistency of financial terminology.