Nvidia's Strategic Bet on Photonics to Power Next-Gen AI

Nvidia commits $4 billion to Lumentum & Coherent for AI photonics supply, while Morgan Stanley names it the top semiconductor stock. The Vera Rubin architecture promises lower AI inference costs.

In a significant move to secure its supply chain for future artificial intelligence hardware, Nvidia has committed $4 billion to long-term supply agreements with photonics specialists Lumentum and Coherent. The chipmaker is allocating $2 billion to each firm, a strategic investment aimed at procuring advanced laser and optical components. This capital will also support research, development, and the expansion of U.S.-based manufacturing facilities, with Lumentum having already announced plans for a new plant.

The core objective behind this substantial investment is the integration of photonics technology to meet escalating AI inference demands while enhancing energy efficiency within massive data centers. Optical solutions are increasingly viewed as critical for overcoming the performance and power limitations of traditional electronics in next-generation computing. Following the deal's announcement, shares in both supplier companies saw notable gains in pre-market trading.

Analyst Sentiment Shifts Following Recent Volatility

Attention is returning to Nvidia's stock after a recent pullback that occurred despite a strong earnings report. Morgan Stanley analyst Joseph Moore has now designated the company's shares as the top pick within the semiconductor sector, displacing Micron from that position. Moore highlighted the stock's attractive valuation, noting it currently trades at approximately 18 times projected earnings for 2027.

Last week, the stock faced pressure even after Nvidia disclosed record-breaking results for the fourth quarter of fiscal 2026. The company reported revenue of $68.1 billion, a 73% year-over-year increase, with its data center segment contributing $62.3 billion of that total. Investor concerns appeared to center on customer concentration, as a significant portion of revenue was attributed to two major hyperscaler clients.

Financial Forecast and the Next Architectural Leap

For the current first quarter of fiscal 2027, Nvidia has provided a revenue outlook of approximately $78 billion. In a related development, CEO Jensen Huang confirmed that initial samples of the new Vera Rubin architecture have already been shipped to customers. Volume production for this next-generation platform is scheduled for the second half of 2026.

A key promise of the Vera Rubin design is a substantial reduction in inference costs compared to the current Blackwell platform. This anticipated improvement in cost efficiency is positioned as a vital factor for clients looking to scale their AI infrastructure economically. The company's dual strategy—securing advanced photonics supply and advancing its GPU architecture—underscores its focus on maintaining a competitive edge in the rapidly evolving AI hardware landscape.