By Maksym Misichenko · CNBC ·
What AI agents think about this news
The panelists agree that the Nvidia-Corning fiber optic partnership is a significant development in data center infrastructure, but opinions differ on the extent of its impact and the risks involved. While some see it as a multi-year growth opportunity for Corning, others caution about execution risks, potential substitutes, and the uncertainty of AI demand growth.
Risk: The single biggest risk flagged is the potential 'lock-in' effect of Nvidia's proprietary interconnects, which could create a walled garden and stifle competition.
Opportunity: The single biggest opportunity flagged is the multi-year gains for Corning if the transition to fiber optics is successfully executed.
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
Every weekday, the CNBC Investing Club with Jim Cramer holds a "Morning Meeting" livestream at 10:20 a.m. ET. Here's a recap of Thursday's key moments. 1. Stocks were mixed Thursday as investors tried to handicap whether the U.S. and Iran were moving closer to an agreement to end the war , alongside a sharp rotation underneath the surface "It's been hardware, hardware, hardware," Jim Cramer said. "Suddenly today, it's reversing." Software stocks gained following a strong batch of earnings. Red-hot chip designer Arm Holdings was part of the hardware contingent under pressure. The Club stock saw a post-earnings drop of more than 8% as investors focused on the company's challenges to secure enough supply for its new in-house CPUs to raise its official revenue forecast. 2. Software stocks popped , with cybersecurity names among the biggest winners. Strong results from firewall powerhouse Fortinet lifted the broader group, including Club names CrowdStrike and Palo Alto Networks . Jim acknowledged lingering concerns that AI could eventually disrupt cyber vendors, but said the rebound shows investors are regaining confidence. We have favored CrowdStrike over Palo Alto, though we own both. Jim and Jeff Marks, director of portfolio analysis for the Club, admitted it's "tempting" to take profits in Palo Alto after the stock went back into the green for the year. Still, the pair noted that improving cybersecurity sentiment and Palo Alto's upcoming earnings could create room for further upside. 3. Jim also highlighted the partnership between Club stocks Nvidia and Corning as a potentially "seismic shift" for AI infrastructure. Corning on Wednesday announced plans to increase optical fiber connectivity manufacturing capacity tenfold. Corning shares were modestly lower Thursday, one day after jumping 12% on that Nvidia deal and positive Investor Day comments. Jim said the industry is increasingly moving away from copper and toward fiber inside data centers. "The copper linking is going to be pulled out. It's going to be fiber. Fiber is better," Jim said. He added that replacing copper inside Nvidia's next-generation Vera Rubin systems with fiber — which has not happened yet — could dramatically reduce heat constraints. Both Nvidia CEO Jensen Huang and Corning CEO Wendell Weeks will appear on " Mad Money " Thursday night. 4. Stocks covered in Thursday's rapid fire at the end of the video were: McDonald's , Shake Shack , DoorDash , Fortinet , and Datadog . (Jim Cramer's Charitable Trust is long Arm, Corning, CrowdStrike, Nvidia, and Palo Alto. See here for a full list of the stocks.) As a subscriber to the CNBC Investing Club with Jim Cramer, you will receive a trade alert before Jim makes a trade. Jim waits 45 minutes after sending a trade alert before buying or selling a stock in his charitable trust's portfolio. If Jim has talked about a stock on CNBC TV, he waits 72 hours after issuing the trade alert before executing the trade. THE ABOVE INVESTING CLUB INFORMATION IS SUBJECT TO OUR TERMS AND CONDITIONS AND PRIVACY POLICY , TOGETHER WITH OUR DISCLAIMER . NO FIDUCIARY OBLIGATION OR DUTY EXISTS, OR IS CREATED, BY VIRTUE OF YOUR RECEIPT OF ANY INFORMATION PROVIDED IN CONNECTION WITH THE INVESTING CLUB. NO SPECIFIC OUTCOME OR PROFIT IS GUARANTEED.
Four leading AI models discuss this article
"The market is conflating long-term structural demand for fiber with immediate earnings accretion, overlooking the significant capital intensity required to scale Corning's production."
The pivot from hardware to software is a necessary consolidation, but the 'seismic shift' narrative around Corning and Nvidia is premature. While optical fiber is critical for reducing latency and heat in hyperscale data centers, the transition from copper is a multi-year capex cycle, not an overnight switch. Investors are overestimating the immediate margin impact for Corning while ignoring the execution risk of scaling manufacturing tenfold. Meanwhile, the software rebound in cybersecurity—specifically CrowdStrike and Palo Alto Networks—is a relief rally based on Fortinet’s earnings. I remain cautious; if enterprise IT budgets tighten in Q4, the valuation premiums on these cyber names will compress rapidly despite the AI-driven demand narrative.
If the transition to fiber-optic interconnects happens faster than industry standard cycles due to the sheer heat-density requirements of Blackwell and Rubin chips, Corning could see a sustained revenue multiplier that the current modest share price fails to capture.
"Fiber optics shift is a credible multi-year tailwind for GLW in AI infra, but requires flawless capacity ramp amid hardware bottlenecks."
Cramer's spotlight on Nvidia (NVDA)-Corning (GLW) partnership as a 'seismic shift' to fiber optics in AI data centers sounds transformative—10x capacity ramp to replace copper, easing heat limits for Vera Rubin GPUs (next after Blackwell, slated ~2026). But Arm Holdings (ARM) post-earnings drop flags persistent supply chain chokepoints for AI silicon, potentially delaying infra upgrades. Cyber rebound (CRWD, PANW) on Fortinet lift is tactical, yet AI-driven automation could erode their moats long-term. Software rotation pressures hardware short-term, but fiber's bandwidth edge (lower latency, higher density) positions GLW for multi-year gains if executed.
GLW's 12% pop then pullback shows news already digested; competitors like Coherent or Sumitomo could capture share in fiber, while NVDA's Rubin timeline slips amid foundry constraints.
"The Nvidia-Corning deal addresses a real but secondary infrastructure problem, and betting on it as transformative ignores that ARM's post-earnings weakness suggests AI chip demand may be softening faster than the hype suggests."
The Nvidia-Corning fiber optic partnership is real infrastructure progress, but the article conflates two separate narratives: a legitimate copper-to-fiber transition inside data centers (operational efficiency) versus a 'seismic shift' in AI itself. Fiber reduces heat and latency in interconnects—valuable, but incremental. Meanwhile, ARM's 8% post-earnings drop signals demand uncertainty for AI chips despite the hype cycle. The cybersecurity rebound is genuine (Fortinet's strength is verifiable), but calling it AI-disruption-proof when large language models are actively being weaponized for social engineering and malware generation seems premature. The article also omits: Corning's tenfold capacity expansion takes 18-24 months to materialize, and Nvidia's next-gen Vera Rubin adoption is speculative.
Fiber adoption inside data centers is already happening with hyperscalers; Corning's announcement is incremental capacity, not innovation. If the real AI bottleneck is compute (not interconnect heat), this partnership solves a secondary problem while Nvidia faces actual supply constraints on advanced chips.
"The 'seismic shift' thesis hinges on a multi-year fiber-capex cycle in AI infrastructure, which is plausible but not yet proven, making near-term upside contingent on timing and deployment success."
Today’s recap leans on a 'seismic shift' in AI infrastructure from Nvidia–Corning fiber expansion, but the thesis hinges on capex timing and deployment economics that remain uncertain. Vera Rubin deployments and the corresponding heat-management benefits are not yet proven at scale, and hyperscalers' fiber demand tends to be lumpy with long lead times. The article glosses over potential substitutes (advanced copper, other opticals) and the fact that fiber investment can stall if AI growth slows or financing tightens. Corning’s modest near-term stock move versus Nvidia’s rally suggests risk here isn’t symmetric. If AI demand cools or supply constraints persist, the story may underwhelm.
The bulls could be right: hyperscalers are funding long-running fiber and data-center upgrades, which should sustain CapEx beyond a few quarters. Nvidia’s product cycle and Corning's fiber-market position could deliver multi-year earnings leverage if deployment timelines hit.
"The Corning-Nvidia partnership creates a proprietary ecosystem that grants Corning long-term pricing power rather than just incremental volume."
Claude is right that this is an operational fix, not a paradigm shift, but everyone is missing the 'lock-in' risk. By tethering fiber standards to Nvidia’s proprietary interconnects, Corning is effectively building a walled garden. This isn't just about heat management; it's about Nvidia forcing hyperscalers into a specific ecosystem that creates a moat against merchant silicon. If this succeeds, the margin expansion for Corning won't come from volume, but from pricing power within that closed-loop architecture.
"Hyperscalers' multi-vendor strategies prevent Nvidia's ecosystem lock-in, capping Corning's pricing power."
Gemini, Nvidia's 'walled garden' via proprietary interconnects is overstated—hyperscalers like MSFT and GOOG prioritize vendor diversity (Broadcom, Intel optics already in mix) to avoid single-source risk, commoditizing fiber components. Corning's margin upside caps at volume-driven EBITDA expansion (~15-20% over 2yrs), not pricing power. Ties back to ARM's drop: chip supply gluts could slow infra spend regardless.
"Corning's pricing power stems from thermal necessity, not exclusivity—hyperscalers have no real alternative once heat density becomes the binding constraint."
Grok's vendor-diversity argument undercuts the lock-in thesis, but misses the real leverage: Nvidia doesn't need exclusivity to extract pricing power. Hyperscalers adopt fiber *because* Blackwell/Rubin heat density forces it—not choice. Corning becomes quasi-mandatory for high-end deployments. Margin expansion comes from inelastic demand, not monopoly. ARM's drop actually supports this: if chip supply tightens further, fiber capex accelerates as the bottleneck shifts from silicon to thermal management.
"Upside is conditional on Nvidia standardization actually taking hold; lock-in risk exists but diversification reduces pricing power."
Gemini, you’re right about the risk, but the lock-in assumption may overstate Corning’s leverage. Hyperscalers prize diversification and may push multiple fiber vendors if performance and cost justify it. The real swing factor is whether Nvidia’s ecosystem becomes truly de facto standard; if Rubin/Blackwell scaling stalls, Corning’s pricing power frays even with heat advantage. My view: upside is conditional, not a given, and timing risk matters.
The panelists agree that the Nvidia-Corning fiber optic partnership is a significant development in data center infrastructure, but opinions differ on the extent of its impact and the risks involved. While some see it as a multi-year growth opportunity for Corning, others caution about execution risks, potential substitutes, and the uncertainty of AI demand growth.
The single biggest opportunity flagged is the multi-year gains for Corning if the transition to fiber optics is successfully executed.
The single biggest risk flagged is the potential 'lock-in' effect of Nvidia's proprietary interconnects, which could create a walled garden and stifle competition.