Nvidia and Corning team up to build three fiber optic factories in the US focused on AI
Nvidia and Corning just announced a partnership that could be a total game changer for artificial intelligence infrastructure in the United States. The two giants are joining forces to build three new factories dedicated to advanced optical technology manufacturing — located in North Carolina and Texas — with a laser focus on Nvidia’s AI systems.
And the market loved the news. 📈
Corning shares surged 14% right after the announcement, while Nvidia stock climbed nearly 3%. The financial terms of the deal were not publicly disclosed, but the scale of the market reaction gives a pretty good sense of how investors view the potential of this partnership.
What is behind this move, though, goes well beyond stock market numbers. We are talking about a multi-year project that will multiply America’s fiber optic production capacity for communications by 10, create at least 3,000 direct jobs, and most importantly, lay the physical foundation for the next generation of AI data centers.
Co-packaged optics: the technology at the heart of this partnership
Although the two companies have not publicly detailed what will be developed at the new facilities, the technical context points clearly in one direction: co-packaged optics. This technology involves integrating optical transmission components directly alongside processing chips, replacing the traditional copper cables that currently connect components inside rack-scale AI systems.
At the GTC 2025 conference, Nvidia CEO Jensen Huang described co-packaged optics as essential for building AI infrastructure. The idea is fairly straightforward to understand: instead of data traveling as electrical signals through copper cables — which generate heat, lose signal strength, and consume more energy as distance increases — they move as photons of light through Corning’s ultra-thin and flexible glass fibers.
Vlad Galabov, an enterprise infrastructure analyst at research firm Omdia, explained the impact of this shift in pretty direct terms. According to him, the light conversion process happens right next to the chip, which means data travels only a few millimeters instead of crossing the entire circuit board. That drastically reduces energy waste. Galabov also pointed out that Nvidia has been pushing the entire ecosystem to innovate faster.
To put it in practical terms, Nvidia’s rack systems, like Vera Rubin, currently use around 5,000 copper cables internally. Replacing those cables with Corning fiber optics would represent a massive leap in energy efficiency, data transfer speed, and the ability to scale systems to hundreds of thousands of GPUs operating together inside a single data center.
Receive the best innovation content in your email.
All the news, tips, trends, and resources you're looking for, delivered to your inbox.
By subscribing to the newsletter, you agree to receive communications from Método Viral. We are committed to always protecting and respecting your privacy.
Why has fiber optics become a top priority for AI?
When we think about artificial intelligence, it is natural to picture powerful chips, cutting-edge GPUs, and language models running on sophisticated servers. But there is a part of the story that often flies under the radar: all that computing power depends on physical infrastructure capable of moving insane amounts of data at ultra-high speeds, with minimal latency and controlled energy consumption. And that is exactly where fiber optics comes in as a centerpiece of this puzzle.
Modern data centers, especially those running large-scale AI workloads, need to connect thousands of GPUs to each other in an extremely efficient way. Traditional copper cables, which still dominate a large portion of these internal connections, have clear limitations in speed, distance, and heat generation. Fiber optics solves all three problems at once: it transmits data through pulses of light, which means less interference, faster speeds, and far less wasted heat.
Corning CEO Wendell Weeks put the issue plainly in a CNBC interview in January: moving photons consumes between 5 and 20 times less energy than moving electrons. For AI systems processing billions of parameters in real time, that difference is not small — it is the kind of advantage that determines whether a data center can scale up or gets stuck in physical bottlenecks.
Fiber optics also allows less signal loss than copper, speeding up reliable communication and reducing the required distance between the hundreds of thousands of GPUs in a data center. As the number of GPUs in a server grows into the hundreds, interconnect distances increase, and that is precisely the scenario where fiber optics becomes far more cost-effective and energy-efficient.
Nvidia figured this out before most people did. Its interconnect systems already point toward a future where optical connections progressively replace electrical cables inside servers themselves and between processing racks. The partnership with Corning is therefore a strategic play to make sure the supply of high-performance fiber optics in the American market keeps up with the speed at which AI demand is growing — and let’s be honest, that demand shows no signs of slowing down.
What does Corning bring to this partnership?
Corning is no ordinary player in this market. Founded 175 years ago, it has over a century of history in advanced materials innovation, and its legacy in optical communications is solid enough to justify the market’s enthusiasm about this announcement. The company invented fiber optics for long-distance communications in 1970 and has since supplied millions of kilometers of cables to connect racks in AI data centers for all the major players in the industry.
Although Corning is widely known for manufacturing all the display glass for Apple iPhones, its optical communications division remains its largest and fastest-growing business segment. The company’s stock has climbed more than 250% over the past year, driven by this century-old company’s rapid pivot toward the new AI economy.
And this partnership with Nvidia is not Corning’s first big move in this space recently. In January, Meta announced it would spend up to $6 billion as the anchor customer to help Corning expand its optical cable factory in Hickory, North Carolina — an expansion expected to create about 1,000 jobs. Now, with Nvidia coming in so aggressively, Corning is solidifying its position as a strategic infrastructure supplier for both sides of the AI ecosystem: the hardware providers and the big tech companies operating the data centers.
Beyond just manufacturing cables, Corning invests heavily in research and development to create fibers with increasingly specific characteristics — lower signal attenuation, higher transmission density, and compatibility with the new optical standards being adopted by hyperscalers and AI hardware manufacturers. Weeks revealed during an exclusive factory tour in January that he had been working with all the different chip companies on how glass will be part of semiconductor packaging going forward.
The broader ecosystem: investments and competition
The partnership between Nvidia and Corning is not happening in a vacuum. It is part of a much broader effort by Nvidia to build a complete optical interconnect ecosystem around its AI systems. In March, Nvidia invested $4 billion in two companies — Coherent and Lumentum — that develop the lasers and components responsible for converting data between light signals and electrical signals. These are precisely the components that work alongside Corning’s optical fibers to make co-packaged optics a reality.
In 2025, Nvidia had also launched two network switches using similar technologies, placing optical components directly next to the primary AI chips. The competition, of course, is not standing still: Broadcom and Marvell have already introduced similar products, while Intel is also developing its own co-packaged optics solutions.
Nvidia, however, has cemented its position in the AI market quite firmly. Its GPUs are critical for the development of large language models and for giving giants like Alphabet and Meta the ability to massively scale their data centers. Nvidia’s stock price has risen roughly 14 times over the past five years, although gains have slowed recently as investors have started diversifying their bets across the broader AI infrastructure ecosystem, including companies like Intel, Micron, and Corning itself.
This context of investment diversification is important for understanding why the partnership announcement hit Corning’s stock so hard. The market recognizes that the AI race is not just about who makes the best chips — it involves an entire supply chain that stretches from semiconductors down to the glass cables connecting everything together.
What changes for data centers and AI infrastructure?
The promise to multiply America’s fiber optic production capacity by 10 is not just an impressive number for a headline. It represents a structural shift in how the United States positions itself in the global race for artificial intelligence supremacy. AI data centers consume enormous quantities of optical cables — both for internal connections between servers and for external network connections that feed these systems with data. With demand for computing capacity growing exponentially, securing a local, stable, and scalable supply chain for fiber optics is practically a strategic imperative for the American tech sector.
From a technical standpoint, the migration to optical interconnects inside data centers delivers concrete benefits that directly impact AI system performance. Reduced energy consumption is one of the most significant. In a scenario where AI data centers already face scrutiny over high energy use, any efficiency gains on this front have a direct impact on sustainability and operating costs for the companies running them. Nvidia, which sells the chips powering these systems, has every incentive to ensure the surrounding infrastructure is equally efficient.
Jensen Huang summed up the moment in an official statement: AI is driving the largest infrastructure buildout of our time — and a once-in-a-generation opportunity to reinvigorate American manufacturing and supply chains. According to him, together with Corning, Nvidia is inventing the future of computing with advanced optical technologies, building the foundations for an AI infrastructure where intelligence moves at the speed of light.
Economic impact and the pride of American manufacturing
Beyond the technology dimension, the creation of at least 3,000 direct jobs extends this partnership past the corporate world, connecting AI expansion with regional economic development in the states involved. North Carolina and Texas are already major tech hubs in the US, and the arrival of advanced manufacturing factories at this level of specialization strengthens these ecosystems even further.
Wendell Weeks, Corning’s CEO, did not hold back his enthusiasm: according to him, what Nvidia is doing is nothing short of extraordinary, not only for the future of AI, but for the American advanced manufacturing workforce. Corning even celebrated the moment by holding an investor day at the New York Stock Exchange, one day before marking its 175th anniversary by ringing the closing bell.
For local workers in North Carolina and Texas, the new factories mean opportunities in a highly technical field with strong long-term growth prospects — after all, the demand for AI infrastructure is not going away anytime soon.
What to expect going forward
The partnership between Nvidia and Corning is a clear signal that the race for leadership in artificial intelligence is not just happening in the realm of software or language models. It also runs, in a very tangible way, through the ability to build and control the physical infrastructure that supports it all. Faster chips, more efficient architectures, and more sophisticated models only make sense if the connectivity backbone supporting them can keep pace — and that is exactly what is being built right now.
This multi-year deal emerged in the same context of the AI investment explosion that kicked off in 2022 with the launch of OpenAI’s ChatGPT, an event that triggered a massive wave of spending on new processors and systems to power cutting-edge AI models and workloads. The Nvidia-Corning partnership is, in a way, the physical manifestation of that revolution: it is not enough to have the best models if there are no cables, factories, and infrastructure capable of sustaining their operation.
In the coming years, as the factories come online and fiber optic production scales up, we are likely to see this technology spread even more rapidly through data center expansion projects across the US and, consequently, influence global AI infrastructure standards. Companies that currently depend on international suppliers for high-performance optical cables may find this new American capacity to be a competitive alternative in both cost and logistics reliability.
Nvidia’s move with Corning also signals a broader trend: major tech companies are shifting from being mere component buyers to becoming strategic partners — and sometimes co-investors — in the supply chain itself. This changes the power dynamics in the industry, brings hardware innovation closer to software innovation, and creates more integrated and resilient ecosystems.
For anyone following the artificial intelligence space closely, this partnership is an important chapter in a story that is only just getting started. The next time you hear about breakthroughs in AI, remember that behind the algorithms and the chips, there is a strand of glass carrying light — and data — at the speed the future demands. 🚀
