Tower Semiconductor announced that it is expanding AI infrastructure deployment through high-performance silicon photonics technology, aiming to support 1.6T data center optical modules that comply with NVIDIA network protocols... Analysts from Yole Group provided insights on this.
“Tower Semiconductor's collaboration with NVIDIA is a strong signal that data center interconnects are entering the next phase of growth. The pace of AI cluster expansion is so rapid that limiting factors are increasingly evolving into the efficiency of data movement between GPUs, switches, and racks.” — Lakshman Srinivasan, Photonics Market and Technology Analyst at Yole Group
This announcement points to a practical industry shift: By combining leading AI network demands with a mature silicon photonics manufacturing base, 1.6T optical modules are getting closer to actual deployment.
The core point is not just about “higher bandwidth,” but also about the readiness of the ecosystem. As optical devices transition from 800G to 1.6T, success depends on repeatable manufacturing, predictable yields, and the ability to scale supply, as this transformation will be driven by real construction projects. In this context, Tower's silicon photonics platform is crucial — it is a proven production environment where many photonic integrated circuit (PIC) teams already know how to validate, industrialize, and mass-produce.
Equally important is that the recent path for 1.6T pluggable modules is expected to remain pragmatic. This means the industry can leverage today's conventional manufacturing technologies to move forward while continuing to innovate in tighter integration over time.
Overall, this partnership is a positive data point for the optical transceiver ecosystem. It further confirms that networks and optical devices have become key drivers of AI performance, and the 1.6T cycle is transitioning from “small-scale ramp-up” to “large-scale deployment,” with the supply chain following suit.
<h2>Tower Semiconductor, Market Value Soars 300%</h2>
This Thursday, Tower Semiconductor's stock price surged again, pushing the Israeli chip manufacturer's market value to over $15 billion.
This move marks a significant increase in Tower's valuation: its market value is now about three times the $5 billion price Intel had promised to pay before abandoning its acquisition two years ago. At that time, Intel ultimately dropped the acquisition plan due to regulatory approval failures. The failure of the deal was widely seen as a setback for Tower. Now, it increasingly looks like a turning point.
In the past six months, Tower's stock price has risen over 160%, with a significant increase this year, making its valuation multiples even surpass NVIDIA. This surge has transformed Tower from a long-overlooked specialty chip manufacturer into one of the most notable beneficiaries in AI infrastructure development
For Russell Ellwanger, CEO of Tower, this transformation is reflected not only on trading screens but also in the production workshop. &#34;When I go home at night, my family, even my neighbors, talk to me about Tower's stock,&#34; he said last month in an interview with Calcalist. &#34;They think I'm a bit special.&#34;
Russell Ellwanger has been at the helm of Tower for twenty years, most of which has been spent in relative obscurity. For years, the company has been viewed as a niche analog chip manufacturer, profitable but not closely tied to the main growth engines of the industry. This perception has changed as data centers are overwhelmed by the pressures of artificial intelligence workloads, forcing a reevaluation of how information is transmitted within servers.
Tower is heavily investing in silicon photonics technology, which replaces traditional copper cable connections with optical transmission. As the amount of data processed by GPU processors increases, copper cables have become a bottleneck in terms of speed and power consumption. In contrast, photonic technology can achieve higher throughput with lower energy consumption, and energy consumption has become one of the most pressing challenges in the field of artificial intelligence.
The company recently announced a $300 million investment to expand its silicon photonics production line, having previously disclosed a $350 million investment earlier this year. Most of the new capacity will be built at Tower's main production base in Migdal HaEmek, northern Israel, which, once completed, will become the company's largest photonics center.
Russell Ellwanger believes the market is just beginning to recognize these impacts. Once the expansion is completed in the first half of 2026, Tower expects its photonics-related business revenue to double, with annual revenue from artificial intelligence-related products alone approaching $1 billion. He estimates that by the end of 2026, 40% to 45% of Tower's revenue will come from data center applications.
This outlook helps explain the market's enthusiasm. Even before the new capacity comes online, Tower has predicted that fourth-quarter revenue will set a record of $440 million and expects full-year revenue in 2025 to reach $1.5 billion, a year-on-year increase of 14%. Profits are expected to approach $200 million, a considerable figure for a hardware manufacturer.
The stock's surge has also reshaped perceptions of Intel's failed acquisition. After the deal fell through, Tower received $350 million in compensation but lost what many viewed as a safe exit opportunity. In hindsight, employees and investors have increasingly divergent views on the outcome. Tower maintained its independence, continued to invest actively during the lengthy regulatory review period, and ultimately secured a solid position in one of the key areas of artificial intelligence infrastructure.
Risk Warning and Disclaimer
The market has risks, and investment requires caution. This article does not constitute personal investment advice and does not take into account the specific investment goals, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article align with their specific circumstances
Invest based on this information at your own risk

Tower Semiconductor collaborates with NVIDIA to expand AI infrastructure using high-performance silicon photonics technology, supporting 1.6T data center optical modules. Yole Group analyst Lakshman Srinivasan pointed out that this collaboration marks a new growth phase for data center interconnects, accelerating the expansion of AI clusters, with data mobility efficiency becoming critical. This announcement reflects industry transformation, as the success of the 1.6T optical module relies on repeatable manufacturing and yield. Tower's silicon photonics platform provides a validated production environment for this, driving the optical transceiver ecosystem towards large-scale deployment

- Tower Semiconductor is expanding its AI infrastructure deployment through advanced silicon photonics technology to support 1.6T data center optical modules compatible with NVIDIA networks.  
- This collaboration highlights a pivotal industry shift towards increased bandwidth and efficient data movement in AI clusters, with expectations for significant revenue growth in photonics-related business attributed to data center applications.  
- Following recent investments and a sharp increase in stock value, Tower's market position has transformed from a niche player to a significant beneficiary in the AI infrastructure sector.

1.6T optical module, Tower collaborates with NVIDIA