AI Advancements Drive 20% Surge in Computing Demand

AI is advancing rapidly, but the computing power required to support these advancements is struggling to keep pace. Models like DeepSeek and ASI-Mini 1 are introducing smarter architectures, which could potentially alleviate the compute crisis. However, these advancements also raise questions about whether they are solving the problem or exacerbating it.

Both DeepSeek and ASI-Mini 1 utilize Mixture of Experts (MoE), an architecture that incorporates multiple expert sub-models. Instead of activating the entire model for every request, MoE selectively activates specialized expert models, reducing computational strain while maintaining performance. This approach enhances compute efficiency, scalability, and specialization, making AI systems more adaptable and resource-conscious. This breakthrough underscores the growing importance of MoE in AI development.

While both models employ MoE, ASI-Mini 1, developed by Fetch.AI, goes a step further by incorporating Mixture of Agents (MoA) or Mixture of Models (MoM). MoA allows multiple autonomous AI agents to collaborate, optimizing resource use and making AI more adaptable. ASI-Mini 1 is also notable for being the world’s first Web3 large language model, excelling in expansion and adaptability.

Optimized compute usage should theoretically reduce overall computing demand. However, Jevons Paradox suggests that efficiency gains often lead to greater adoption, ultimately driving demand even higher. DeepSeek’s ability to deliver high-performance AI at lower costs is a prime example—by making AI more accessible, it fuels greater investment in AI projects, intensifying the need for infrastructure. As a result, the focus shifts toward ensuring solutions are not only cost-efficient but also scalable and adaptable to sustain AI’s rapid growth.

Both large language models (LLMs) and AI agents are intensifying the demand for computing power, requiring substantial resources for training, inference, and real-time decision-making. LLMs, particularly the latest iterations with billions of parameters, are computationally expensive not just during training but also during inference, where generating responses at scale remains resource-intensive. AI agents, operating in dynamic environments, introduce continuous workloads, constantly analyzing incoming data and making autonomous decisions in real time. This sustained computational demand places additional strain on infrastructure, requiring consistent access to high-performance compute resources.

GPUs remain the foundation of AI infrastructure, but their high costs, supply chain constraints, and availability pose significant challenges for businesses scaling AI operations. This surge in AI adoption makes high-performance, cost-efficient, and scalable infrastructure an imperative, particularly as businesses seek flexible, transparent, and globally distributed compute solutions to maintain a competitive edge.

The market is experiencing an infrastructural shift, where companies must rethink how they build, deploy, and sustain AI systems. AI applications are no longer limited to research labs or enterprise automation; they are embedding themselves into consumer products, financial systems, and real-time decision-making engines. AI agents, once a niche concept, are now being deployed in autonomous trading, customer interactions, creative fields, and decentralized networks, all of which require constant, real-time compute power.

There is also an evolution in how AI infrastructure is funded and scaled. Businesses are not just developing better models—they are strategizing around compute access itself. The scarcity of GPUs, the need for decentralized compute solutions, and the rising costs of cloud AI infrastructure are becoming as critical as AI model improvements themselves. Companies that once focused solely on AI capabilities now must navigate compute economics just as carefully. Those who fail to plan for infrastructure growth risk being left behind.

In summary, while DeepSeek and ASI-Mini 1 introduce innovative architectures that enhance compute efficiency, they also highlight the growing demand for computing power in AI. The focus must now shift toward ensuring that AI infrastructure is not only cost-efficient but also scalable and adaptable to sustain the rapid growth of AI. Companies must strategize around compute access and plan for infrastructure growth to keep pace with the accelerating demand for AI computing power.