NVIDIA Commits $6 Billion to Build a U.S. Open-Weight AI Alternative to Chinese Models
NVIDIA has spent the last few years becoming one of the biggest winners of the artificial intelligence boom. Now, the company appears ready to take a bigger role in the technology that runs on its chips.
NVIDIA is reportedly committing $6 billion to license technology from AI startup Poolside, with more than 100 Poolside employees expected to join the company. NVIDIA is also investing another $1 billion in Poolside, at a reported pre-money valuation of $12 billion.
The move will strengthen NVIDIA's work on Nemotron, its family of open-weight AI models, and comes as Chinese AI companies continue to make progress with models that are attracting developers and businesses around the world.
From Chips to AI Models
NVIDIA's biggest advantage has traditionally been its hardware.
Its GPUs have become essential for companies training and running advanced AI systems, turning NVIDIA into one of the most valuable companies in the technology industry.
But the AI business is changing quickly.
The biggest players are now competing across the entire technology stack — from chips and data centers to software and AI models. NVIDIA appears to be making sure it doesn't remain only the company supplying the hardware.
The Poolside deal gives it access to both technology and experienced AI engineers, potentially speeding up its efforts to build more capable models of its own.
Why Poolside?
Poolside is a relatively young AI company that has focused heavily on AI systems for software development and other advanced applications.
Its technology and engineering talent are the main attractions for NVIDIA.
More than 100 Poolside employees are expected to move to NVIDIA, where they will work on the company's Nemotron AI efforts.
Rather than a straightforward acquisition, the arrangement combines a technology-licensing deal with a separate investment and the transfer of a significant portion of Poolside's team.
For NVIDIA, that could be a faster way to add expertise than building an equally large team from scratch.
The Growing Importance of Open-Weight AI
One reason this move matters is the growing popularity of open-weight AI models.
Unlike completely closed systems, open-weight models can give developers more freedom to run and adapt AI technology for their own needs, depending on the licensing terms.
That makes them attractive to companies that want more control over their data and AI systems. Governments and organizations developing their own national or industry-specific AI capabilities are also showing greater interest in this approach.
The market has changed considerably over the past year as several open models have demonstrated that they can compete with much more expensive proprietary systems in specific areas.
China's Role in the AI Race
Chinese AI companies have played an important part in driving that change.
Companies such as DeepSeek have drawn global attention with models that emphasize efficiency and open access. Other Chinese developers are also working aggressively to improve their AI capabilities.
That has added another layer to the competition between the U.S. and China.
The question is no longer simply who can build the biggest AI model. Cost, efficiency, accessibility and the ability to customize a model are becoming just as important.
For U.S. technology companies, developing competitive open-weight models has therefore become a strategic priority.
NVIDIA's latest move appears to be part of that broader shift.
Nemotron Gets a Bigger Team
NVIDIA has been developing its Nemotron family as part of its open AI strategy.
Bringing in Poolside's engineers could give the project additional momentum, particularly as NVIDIA looks to build models that can compete with leading systems from both the U.S. and China.
The company also has an advantage that many AI startups do not: access to enormous amounts of computing power and a huge ecosystem of developers already using NVIDIA technology.
If NVIDIA can combine that infrastructure advantage with competitive AI models, it could become a much more influential player in the model market.
Could NVIDIA End Up Competing With Its Own Customers?
There is an interesting complication to NVIDIA's strategy.
Some of the world's biggest AI companies — including OpenAI and Anthropic — are also major users of NVIDIA's chips.
NVIDIA has benefited enormously from selling computing power to these companies, regardless of which model ultimately wins.
Moving deeper into AI models could change that relationship.
The company would increasingly be both a supplier and a competitor.
That doesn't necessarily mean NVIDIA will directly challenge every major AI laboratory, but it does show that the company wants more control over what happens above the hardware layer.
A Bigger AI Strategy
The Poolside deal is also part of a much broader expansion by NVIDIA.
The company is increasingly involved in AI infrastructure, software, startups and financing alongside its core semiconductor business.
That strategy makes sense as AI becomes a full technology ecosystem rather than a single product category.
For NVIDIA, the opportunity is enormous.
If AI continues to expand into everything from enterprise software and robotics to healthcare, autonomous vehicles and data centers, demand will grow across the entire technology stack.
Having a presence at several levels could make NVIDIA less dependent on any single part of that market.
What It Means for the U.S.-China AI Competition
The deal also highlights how closely technology and geopolitics have become linked.
The U.S. has major strengths in AI research, semiconductor design, computing infrastructure and venture capital. China, meanwhile, is investing heavily in domestic AI development and building models that are increasingly competitive.
Open-weight AI is becoming one of the areas where the two ecosystems are competing most visibly.
By putting billions behind Poolside's technology and talent, NVIDIA is signaling that it wants the U.S. to remain competitive in this part of the market.
The Road Ahead
The real test for NVIDIA will be what comes next.
Spending billions and hiring experienced engineers does not automatically guarantee that Nemotron will become a leading AI model.
The models will ultimately have to perform well, attract developers and prove useful to businesses.
But NVIDIA has several advantages: a huge developer community, access to computing infrastructure and deep relationships across the AI industry.
If those strengths can be combined with Poolside's technology and talent, NVIDIA could significantly strengthen its position in the AI model market.
Key Takeaways
$6 billion: Reported NVIDIA commitment to license Poolside's technology.
$1 billion: Separate NVIDIA investment in Poolside.
100+ employees: Poolside engineers and other staff expected to join NVIDIA.
$12 billion: Reported pre-money valuation for Poolside.
Nemotron: NVIDIA's open-weight AI model initiative.
Main battleground: Competition between U.S. and Chinese AI developers.
The Bottom Line
NVIDIA's latest move shows just how quickly its role in AI is expanding.
The company built its dominance by supplying the chips behind the AI boom. Now it wants a bigger say in the models and software running on those chips.
The $6 billion Poolside deal could give NVIDIA a significant boost in that effort, while also strengthening the U.S. push for competitive open-weight AI.
For NVIDIA, the message is becoming increasingly clear: the future of AI isn't just about who makes the fastest chips — it's also about who builds the technology running on them.
