Glass – AI’s unsung hero

Glass is everywhere, shaping our lives in ways we often overlook. From drinking glasses and light bulbs to the windshield on your car and the screen on your smartphone, glass is a vital part of daily life.

 

But did you know that glass also plays a critical role in powering some of the most advanced technologies in the world? Artificial intelligence (AI) – the powerhouse behind tools such as ChatGPT, autonomous vehicles, fraud detection systems and virtual assistants – wouldn’t be possible without glass. The truth is, AI’s existence relies on revolutionary glass technology.

 

The backbone of AI: fibre in data centres 

 

Every AI application starts its journey in a data centre – the birthplace of AI. Essentially, a data centre is a facility that houses rows and rows of powerful computers designed to store, process and manage vast amounts of information. These centres can operate two main networks: a front-end network that supports processing of user/client applications, such as the files in your phone or laptop; and another backend network that powers advanced AI algorithms.

 

The backend network is filled with server racks populated with powerful chips called graphics processing units (GPUs), interconnected with fibre connections designed to support a range of AI applications (machine learning or neural networks, for example) optical fibre – glass thinner than a human hair – is the foundation of these neural networks. It is responsible for transmitting massive amounts of data at lightning speeds. 

 

The first low-loss optical fibre was created by Corning Incorporated back in 1970. Today, optical fibre connects complex networks and allows AI to process and analyse enormous datasets. Without fibre, data centres couldn’t compute even the most basic applications, let alone perform AI tasks ranging from understanding human language to generating lifelike images. 

 

AI needs fiber – a lot of it 

 

The rise of AI has created a demand for optical fibre like never before. AI networks require more than ten times more fibre than typical frontend data centre networks, driving a need for smaller, streamlined, more powerful fibre components. Corning, a global leader in cutting-edge glass solutions, has prepared for the AI era.

 

Over the past several years, Corning’s engineers and product development teams have worked diligently to reinvent several key connectivity products for AI networks – making componentry both smaller, to optimise existing space within data centres, and faster and easier for installers to deploy. Corning’s optical fibre technology is the backbone of the world’s largest AI networks, enabling faster and more efficient computations for your favourite apps and services.

 

Beyond optical fibre, Corning also produces glass components essential for the semiconductor industry, which powers the chips and processors that make AI calculations possible. By providing the advanced materials AI systems depend on, Corning is moving the world forward with every connection.

 

Meeting AI’s growing demands 

 

AI is advancing rapidly, and behind every new application – whether it’s a smarter virtual assistant or a self-driving car – there’s an increasingly complex network making it possible. These systems require constant innovation and collaboration to keep up with growing demand. Companies such as Corning are leading the charge, working tirelessly to develop solutions that ensure AI systems can keep up with the pace of technological progress. 

 

For example, Corning® SMF-28® Contour optical fibre has been integrated into our latest outdoor cable offering, the Contour™ Flow cable – designed to connect AI campuses over long distances and enable seamless data transmission. We have seen a high adoption of this solution. One notable example of this can be seen in Lumen Technologies and Corning’s August 2024 agreement to supply Lumen Technologies with a substantial supply of this next-generation optical cable. This fibre-dense cable will more than double Lumen’s US intercity fibre miles, offering significant capacity to major cloud data centres racing to stay ahead of AI workloads and high bandwidth applications fuelled by massive amounts of data.

 

Looking towards the future 

 

Advancing AI capabilities isn’t a solo effort – it takes cross-industry collaboration to make these systems a reality. Corning is working with some of the biggest names in technology to push the boundaries of what’s possible.

Most recently, Corning was named one of NVIDIA’s technology innovation collaborators in its silicon photonics ecosystems, which will significantly increase AI reliability while saving energy and network operational expenses.

Additionally, Corning and Microsoft are collaborating on a new transformative program that will help shape the future of AI and cloud networks around the world, a strategic manufacturing collaboration to accelerate the production of Microsoft’s Hollow Core Fiber (HCF). Corning’s fibre and cable manufacturing facilities in North Carolina will produce Microsoft’s fibre, as Microsoft seeks to advance the performance and reliability of Azure’s Cloud and AI workloads.

 

These collaborative efforts show that innovation doesn’t happen in isolation – it requires teamwork and shared expertise across industries. By developing groundbreaking glass technologies and working with industry leaders, Corning is ensuring the world stays connected and that AI continues to evolve to improve lives. With every new connection, Corning is helping to shape the future – one strand of fibre at a time.

---

Michael Crook is a Data Centre Market Development Manager. He supports our hyperscale, multi-tenant, and enterprise customers with new fibre-optic innovations and commercial solutions. With over 15 years of experience, Michael has amassed a great deal of knowledge in designing and building fibre-optic network infrastructures for data centre and carrier environments.