NVIDIA places a $40 billion bet on open source, profiting from both selling "shovels" and mining gold at the same time.

Expected??

Nvidia was recently revealed to have bet $6 billion on open-source models.

On the table, Nvidia clinks glasses with several closed-source large model companies, saying "OpenAI, Anthropic, it's been a pleasure cooperating with you!"

Under the table, it quietly pulled out $6 billion and handed it to Poolside, a startup founded by the former CTO of GitHub: "Go quickly and build a world-class open-source large model that can rival the top-tier global models!"

Moreover, this deal just finalized this week also targets strong open-source players such as DeepSeek and Kimi K3.

Nowadays, open-source models have generated a considerable share of AI-generated tokens, and Jensen Huang hopes this proportion will continue to rise, to continuously stimulate demand for its own chips.

Some analysts believe that "Nvidia doesn't care which AI model wins in the end, as long as the world keeps producing more and more models nonstop..."

Nvidia's Grand Ambition and Open-Source Layout

Last month, Jensen Huang broke his usual low profile and posted his first post ever on X:

This post was co-signed by more than 20 companies, advocating that if the U.S. AI industry wants to achieve development, it must attach equal importance to both closed-source and open-source models.

Now, Nvidia is stepping in personally to fulfill this glorious mission.

Recently, Nvidia invested $1 billion in AI startup Poolside, and additionally paid $6 billion to purchase the non-exclusive right to use its "model factory" technology, hoping to recruit all more than 100 engineers of the company.

The general public widely believes that Nvidia's move is a big bet, it wants to leverage Poolside's technology and team to challenge the world's most powerful open-source models.

This move is not abrupt at all.

In the past few years, leading AI labs such as OpenAI, Anthropic and DeepMind have invested major resources in closed-source model R&D, and none of them have open-sourced their most advanced models.

This also leaves development space for models like DeepSeek and Kimi K3.

Facing this landscape, Nvidia seized the opportunity and took the completely opposite approach.

Since the end of 2025, Nvidia has successively released the Nemotron 3 series of models, among which the largest Ultra version once became the most powerful open-weight model in the United States in June this year.

Moreover, Nemotron's open-source attitude is completely like a kind of declaration: apart from model weights, no other data is hidden.

Training data, training recipes, post-training methodologies and GPU cluster training software are all made public.

In addition to independent R&D, Nvidia also formed a small open-source group in March this year — the Nemotron Alliance.

Its 8 members are all outstanding players with strong backgrounds, including Mistral, Perplexity, Thinking Machines Lab, Cursor, LangChain, Reflection AI, Black Forest Labs, Sarvam...

As the leader, Nvidia provides DGX Cloud computing resources, while each member company contributes its own technology and data to jointly train an open-source model, which will serve as the foundation for the next-generation Nemotron 4 series.

Besides, Nvidia's open-source layout is not limited to language models.

In the robotics field, Nvidia has released the Isaac GR00T series of models; in the physical world simulation field, there is the Cosmos series; there are also the Alpamayo series for autonomous driving and the Clara platform for the biomedical field...

How a Company Founded by Idealists Became a Piece of Nvidia's Puzzle

Poolside, which literally means beside the swimming pool.

It was founded in 2023 by software developer Eiso Kant and Jason Warner, former CTO of GitHub.

It is said that the name came from a financing negotiation between the two founders and a large tech firm, when the executive of the other party proposed to hold a relaxed "poolside informal meeting".

It is unknown whether the negotiation was successfully closed in the end, but the two founders thought the name was excellent the moment they heard it: easy to remember and full of fun.

Last October, Poolside announced that it would build a 2 GW data center in Texas. Unexpectedly, in April this year, the project partner withdrew, and the subsequent $2 billion financing also fell through.

The dual pressure of capital and computing power once pushed Poolside to the verge of operational difficulties.

What supported them to continuously develop models with limited resources is their internal "model factory" system.

According to Eiso Kant, his company usually takes 5 to 8 weeks to train a model until release, the R&D team is no larger than 70 people, and it can run 10,000 to 20,000 experiments per month.

This capability is exactly what Nvidia values.

After completing this cooperation with Nvidia, the remaining team of Poolside will mainly consist of three people: the two founders and the operation executive.

It is worth mentioning that Eiso Kant, the founder of Poolside, once said in a podcast program:

I want to see a world where there are 100 foundation model companies, not just 5, even if we could have been one of those 5.

One More Thing

In fact, the transaction structure for Poolside this time is not the first one Nvidia has adopted.

Non-exclusive licensing, recruitment of core talents, and the original company retaining the name of independent operation — none of these constitute a direct acquisition, so it does not need to go through antitrust review.

Analyst Stacy Rasgon commented that "doing so may allow the narrative that 'competition still exists' to continue to be maintained."

Nvidia hopes to firmly grasp all the key nodes of the entire open-source ecosystem in its own hands.

Nvidia's vice president of Generative AI Software once said:

Models are just by-products, not our core business.

The vice president of Applied Deep Learning Research also once said that the primary purpose of developing the Nemotron series is to ensure the continued existence of Nvidia.

After all, as the effect of Moore's Law weakens, only when the entire AI ecosystem continues to expand and the number of entities developing and applying models keeps growing, can Nvidia's demand for computing power keep rising.

This article is from the WeChat Official Account "QbitAI", written by Cheng Qian, authorized for release by 36Kr.