Hyper Connectivity and the GenAI Boom

There have been industry booms throughout history. Growing up in California, the first one I remember learning about was the Gold Rush. It was fascinating to learn how entire towns and economies grew around mines and miners. As an adult, looking back on those history lessons, I can now reflect with more context and understand how it relates to more recent history, such as the Internet boom, the housing boom, and the oil boom in Canada. In each of these booms, there was incredible growth in the verticals and markets directly related to them, as well as the markets around them. Infrastructure played a crucial role in supporting these booms, whether it was in telecom, roads, towns, or the materials needed for construction. Enter Generative AI.

Artificial Intelligence is not a new concept. In fact, you can trace its roots back to the dawn of the computer age with the Turing Test. In 1950, Alan Turing devised a test to determine "a machine's ability to exhibit intelligent behavior equivalent to, or indistinguishable from, that of a human." AI and Machine Learning have been in use across various industries, leading to a boom in the Data Center world. Ethernet switches have grown in radix and speed per lane, servers have become faster and more efficient in processing data, and power requirements for computing have skyrocketed. The land and resource requirements for Data Centers have grown at a historic rate, supporting the giants of the Cloud Compute era.

Generative AI is a new breed of AI that brings capabilities to a much larger audience. It allows for the creation of new things based on prompts fed into large language models, unlike previous AI systems that could only find patterns in random feeds of data and information. As more people and industries adopt Generative AI, we can expect an increased demand for the infrastructure that supports it. This demand will be driven not only by user-generated traffic and interaction with AI but also by the data created by IoT devices, smart buildings, and smart cities.

Several industries will have the opportunity to take advantage of this modern-day AI Gold Rush. Some of the obvious ones have already made headlines, with market valuations exceeding one billion dollars. However, there are other industries that will also have the opportunity to grow with the adoption of Generative AI.

Compute: While 英伟达 has been at the forefront of this field, there are other players in the market of High-Performance Compute. Ethernet chip makers like 博通 will provide solutions to compete for market share. Servers equipped with accelerators and other specialized hardware to support AI workloads, along with software for orchestration, will be in high demand.

Construction: Companies like Mortenson and Turner Construction Company are already well-positioned in the Data Center construction market. This market is estimated to be over 200 billion USD in size in 2023, with a projected Compound Annual Growth Rate (CAGR) of 7%. Just as important as the compute infrastructure will be the facilities that house it. Acquiring land, providing designs for power, cooling, and security, and constructing facilities capable of supporting the compute infrastructure will be vital.

Connectivity: All the High-Performance Compute and power in the world will not enable AI adoption without proper connectivity. This includes connectivity within the massive data centers that house AI hardware, as well as connectivity that links these data centers to service providers and cloud service providers. Additionally, connectivity is crucial for users to create prompts and leverage the capabilities of Generative AI. Recent initiatives like President Biden's #broadbandforall have highlighted the importance of connectivity, with companies like 康宁公司 , CommScope , and Prysmian contributing to the expansion of fiber optic networks. Connectivity will continue to be a critical component for the adoption of not just Generative AI but all data-driven applications.

The adoption of Generative AI will drive the need for more infrastructure to support its continued use. From home internet connections to service providers building new or updated networks, from manufacturers of components for High-Performance Compute and the networks to support them to the companies involved in constructing the facilities that house and protect data—Generative AI will enable many companies to thrive in our hyper-connected world. It will also act as a catalyst for the supply and construction industries, driving the demand for components, networks, and facilities.