AI capital expenditure is "overflowing" from GPUs to the entire industry chain! Corning (GLW.US) and Qualcomm (QCOM.US) secure major contracts on the same day, strong endorsement for data center investment confidence

According to Zhitong Finance APP, as the AI capital expenditure boom spreads from Nvidia’s GPUs to a broader industrial chain, two vastly different companies—glass manufacturer Corning (GLW.US) and mobile chip giant Qualcomm (QCOM.US)—announced two heavyweight deals on the same day that are redefining the boundaries of AI infrastructure investment.

On Tuesday (September 8), Corning announced a multi-billion-dollar fiber supply agreement with Verizon, lasting until 2032; Qualcomm disclosed that it had issued warrants worth $4 billion to Amazon, as part of AWS’s server chip procurement agreement valued at up to $60 billion. Together, these deals send a clear signal: AI data center capital expenditure is spilling over from the “chip layer” to the “network transmission layer,” and the list of beneficiaries is rapidly expanding.

Corning: From Glass Manufacturer to Key Player in AI Optical Networks

The agreement between Corning and Verizon covers over 80 million miles of high-density optical fiber and connectivity solutions between 2027 and 2032. Verizon will use this order to advance two major strategies: first, expanding fiber broadband coverage to 40–50 million access points in households; second, building long-haul backbone networks connecting AI hyperscale data centers. Verizon Business CEO Kyle Malady stated that the company is deploying a "converged architecture" unifying mobile and broadband connectivity into a seamless experience, while building high-capacity, low-latency backbone networks required by AI hyperscale customers.

This marks Corning’s fourth major long-term AI infrastructure order recently. Previously, Corning secured a multi-year fiber supply agreement with Meta worth up to $6 billion; partnered with Nvidia to expand U.S. optical connectivity manufacturing capacity, with Nvidia committing up to $3.2 billion to build three new fiber manufacturing plants; and signed a multi-billion-dollar data center fiber supply deal with Amazon. Corning CEO Wendell Weeks previously revealed that deals with two unnamed hyperscale customers are even "larger" in scale than Meta’s $6 billion agreement.

Behind these orders is a key statistic: a single AI data center node requires around 16 times more fiber than a traditional switch. From cloud providers to chip companies, and now telecom operators, AI’s demand for fiber is driving an unprecedented structural growth cycle. Boosted by this, Corning’s stock price rose 8% on Tuesday, with a cumulative gain of 90% this year.

Qualcomm: The Critical Leap From Mobile Chips to AWS Custom Chips

If Corning’s story is “AI demand radiating from chips to the periphery,” then Qualcomm’s story is about “a mobile chip company trying to break into AI data centers.”

According to regulatory filings from Qualcomm, the company has issued warrants to Amazon, allowing AWS to purchase up to 25 million shares of Qualcomm stock at $161.26 per share, with the potential stake worth about $4 billion. The warrants will vest in tranches based on Amazon’s order volume with Qualcomm, reaching a potential total of $60 billion. This deal is part of Qualcomm and Amazon’s "multi-generational" custom chip collaboration, where the two will jointly develop custom chips for AI inference, and co-develop optical interconnect solutions with speeds up to 1.6T and beyond.

Qualcomm CFO Akash Palkhiwala confirmed at a Goldman Sachs conference on Tuesday that revenue from chip manufacturing for Amazon will begin in the December quarter and will be a "core component" of the company’s $15 billion data center revenue goal for fiscal 2029. He also revealed that Qualcomm "is collaborating in a similar way with another data center customer"—hinting that Qualcomm’s data center customer list is expanding beyond Meta.

In June of this year, Qualcomm released the Dragonfly C1000 processor for data centers, targeting agent-based AI workloads. Bank of America forecasts that the global CPU market could more than double from $27 billion in 2025 to $60 billion in 2030. Qualcomm’s transformation is a macro bet on the repricing of CPU demand in the AI inference era.

The Spillover Effect of AI Infrastructure Investment: From Nvidia to the Whole Tech Supply Chain

The simultaneous announcement of these two deals is no coincidence. They both point to an accelerating structural trend: the beneficiaries of AI data center spending are now proliferating from GPU manufacturers across the broader infrastructure supply chain.

In after-hours trading on Tuesday, Intel (INTC.US) and AMD (AMD.US) rose 9% and 6% respectively, Hewlett Packard Enterprise (HPE.US) was up 8%, and photonics company Coherent (COHR.US) added 7%. Year to date, Hewlett Packard Enterprise and AMD have both doubled their market caps, while Intel has nearly tripled. On the same day, Goldman Sachs raised its outlook for the optical module industry, and Coherent closed up 7.24%. Deutsche Bank initiated coverage of the AI hardware sector, naming Coherent and Lumentum as its “highest conviction” AI hardware picks.

The macro background to this spillover trend is that AI capital expenditure is shifting from “whether to invest” to “invest as much as possible.” The Big Four U.S. CSPs (Microsoft, Google, Amazon, Meta) saw Q2 capital expenditures jump 86% year-on-year, with the total expected to surpass $886.7 billion in 2026. TrendForce has raised its combined capex estimate for the top five U.S. CSPs plus four top Chinese CSPs to about $830 billion.

Concerns Amidst the Boom: 71% of Americans Oppose Data Centers

However, the feverish expansion of AI infrastructure is not without cost. According to a Gallup poll in May 2026, 71% of Americans oppose the construction of data centers in their communities. This social resistance has directly translated into financial risk—reportedly, Anthropic is preparing to list “negative public perception of AI and data centers” as a risk factor in its upcoming IPO prospectus.

At the same time, persistently rising global bond yields are putting tangible pressure on the financing costs of AI capital expenditure. The 10-year U.S. Treasury yield has risen to around 4.81%, with the 30-year approaching 5.25%, meaning that the debt financing costs for future AI infrastructure are systematically rising.

Conclusion

The two deals announced on the same day by Corning and Qualcomm provide the latest evidence that AI infrastructure investment is still spreading at an accelerating pace. From glass fiber to custom chips, from telecom operators to cloud giants, the beneficiary chain of AI capital expenditure is extending with unprecedented speed. As Corning’s CEO put it, this 175-year-old glass company is "leading the AI revolution at the speed of light." But with 71% public opposition and continuously climbing financing costs, the market is also being reminded that this AI infrastructure feast is not without limits.