Despite the recent decline in bitcoin prices, there is a growing trend among larger mining operations to explore opportunities in AI. While most bitcoin-only miners are not pivoting to AI, larger mining companies are leveraging their existing infrastructure to expand into artificial intelligence.
Bitcoin miners typically use specialized ASICs chips that are designed for bitcoin mining and cannot be repurposed for other tasks. However, the infrastructure that these companies have built, such as cooling systems, security, and access to cheap energy, can be utilized for AI projects.
Texas-based crypto miners, Applied Digital and Iris Energy, have recently announced expansions into AI. Applied Digital secured a $460 million deal to host AI cloud computing in its data center, and Iris Energy announced a revitalization of its high-performance computing data center strategy, with a focus on AI.
The motivation behind these expansions is to reduce reliance on bitcoin and create diversified revenue streams. By offering AI computing services, mining companies are less vulnerable to fluctuations in bitcoin prices. This strategic move allows them to position themselves favorably in the market and act strategically compared to their peers.
Hut 8, a cryptocurrency mining company, started its pivot towards AI early by investing in data centers and cloud regions that could be devoted to high-performance computing. They utilize Nvidia GPUs that can handle a wider range of workloads, including AI and machine learning. Hut 8 believes in the convergence of mining, data centers, and AI, and they are beginning to see this vision become a reality.
Another aspect driving the expansion into AI is the pivot from Ethereum mining to AI. When Ethereum transitioned to Proof of Stake, rendering ETH-mining equipment obsolete, the more versatile GPUs that were used for Ethereum mining were repurposed for AI workloads.
One real-world example of how these mining chips power AI development is the case of XYZ AI, a startup that uses the computing power from companies like Hut 8 to train their models and convert plain text into 3D imagery. This demand for processing power is not limited to generative AI; it extends to medical technology, gaming, biology, CAD drawing, and many other fields.
While the expansion into AI is still in its early stages for many mining companies, it is seen as a viable way to diversify their operations and reduce systemic risk. The global chip shortage has accelerated this trend, as mining companies with existing infrastructure and resources can leverage their capabilities to enter the AI market.