The advancements in generative AI have brought about significant changes in the way people live and work. From developing legal strategies to augmenting workflows and enabling self-driving cars, the impact of language models, image diffusion models, and computer vision is undeniable.

However, the primary challenge in scaling these systems is the availability of compute resources. The demand for chip production has far exceeded the supply, resulting in increasing wait times and hourly rates for spot instances of popular chips like Nvidia’s A100 and H100. This shortage is caused by a combination of factors such as materials constraints, supply chain disruptions, surging demand, geopolitical tensions, and long production cycles. Even essential materials like advanced silicon, specialized substrates, and memory chips are facing shortages.

In 2023, datacenter revenues for AI workloads reached approximately $100 billion. However, setting up and operating data centers require significant upfront capital investment, and external financing is often needed. Despite the decreasing price of computational performance, the demand for AI-specific compute is doubling every six months, making it challenging to keep up with the supply. This shortage presents a massive opportunity for investors, as the demand for compute is expected to increase exponentially in the coming years.

To address the compute shortage problem, a new form of crypto network called “Decentralized Physical Infrastructure Networks” (DePINs) is emerging. It is estimated that there are 1.5 billion consumer GPUs and six million datacenter GPUs globally, which could potentially be utilized to meet the increasing demand for compute. These decentralized networks incentivize GPU owners to contribute their resources in exchange for rewards and create a decentralized networking layer to make these disparate GPUs usable for AI developers. This also offers a more affordable alternative to traditional cloud providers, with costs up to 90% cheaper.

In the coming year, DePINs are expected to play a significant role in the AI race. With the current shortage of affordable GPUs, these decentralized networks offer a solution to distribute AI workloads across a previously untapped pool of qualified hardware. GPUs are the currency of AI, and DePINs are here to provide it.