By building around a model whereby users must find the most energy-efficient sources to unearth new assets, bitcoin mining helped to create a new system of computing. Linking a range of computing systems, decentralized physical infrastructure networks (DePIN) reduce processing costs as well as risk. These complex computing networks are also finding new use cases in areas like artificial intelligence and tokenization.
A new report from Javelin Strategy & Research, Decentralized Infrastructure and Computation: How Does It Compare?, looks at how DePIN offers a cost-effective approach by being able to truly scale and make the network stronger. The benefits of decentralized computation and infrastructure networks outweigh other computing methods in use today, according to Joel Hugentobler, Cryptocurrency Analyst at Javelin and the author of the study. But he warns that this technology is still immature, so businesses must be careful and consider every aspect before moving to implement it.
Spreading the Risk Around
The logic behind DePIN is simple: Rather than operating a single network system in a single place, businesses run multiple servers in multiple areas. By force of habit, many multinational companies often concentrate their computing processes in a single location.
“Even a business as large as Facebook has centralized data servers in one or two locations,” Hugentobler said. “They have all their computation power localized. With all the tech companies located in Silicon Valley and other parts of California, if something like an earthquake were to happen, that would create huge havoc on their business operations. We could go on another 50 or 100 years and there not be an earthquake, but at the same time, it also could happen a week from now. Nobody knows.”
The decentralized architecture of DePIN essentially eliminates the risks of service disruptions by spreading the computational resources around the globe. In the event a node or even several nodes failing, the remaining network components maintain service. What’s more, by leveraging blockchain technology and offering access via cryptographically secure access and storage, DePIN safeguards stored data against unauthorized access or manipulation by third parties.
In addition to eliminating centralized risk, the DePIN model provides lower overhead costs. Managing resources for infrastructure often requires providers to make expensive long-term bets on building out capacity. DePIN keeps that infrastructure delivery within the principles of a market economy, maintaining a natural balance of supply and demand. Participants in the network can expand as demand dictates.
In addition, participants contributing to a network bear the cost of maintaining the equipment. That significantly reduces the overall operating costs for users as opposed to building their own infrastructure.
A Model Arising from Mining
Bitcoin mining paved the way for decentralized models. Bitcoin mining started out on simple computers and even laptops. But it’s built into the model that the difficulty of creating new assets—and the necessity for stronger computing enterprises—increases every four years.
As a result, mining organizations began a somewhat centralized model, but they offered incentives to participants to provide the additional computational resources they needed. Over time, the model became truly decentralized.
When the model matured, innovators developed specialized mining equipment to help them solve the mining algorithm over time. The systems are much more efficient now, in part because they led innovators to seek creative ways to gain low-cost electricity.
Centralized systems require 100 megawatt-hours per day, and sometimes more, because of their reliance on large, single data centers. Decentralized systems, on the other hand, optimize processing efficiency and consume less than half of that per day, making them more energy-efficient in many cases.
Energy efficiency also has an impact on operational costs tied to the centralized data centers, which can average around $1 million per month. Decentralized systems’ operational costs are a fraction of that. They also provide ancillary benefits.
“Bitcoin miners have gone to very small towns in Africa where there’s a small waterfall,” Hugentobler said. “They’ll hook up a hydropower Infrastructure, so these towns that didn’t have access to electricity before now have this electrical infrastructure to tap into.”
In Texas, bitcoin mining operations are fairly common, as is oil drilling. When the grid needs additional power for oil drillers, the bitcoin miners contractually shut down and provide that extra load to the grid.
Decentralized computation can use this same model for electrical grids all around the world and provide resiliency. Even when a handful of grids are shut off, the network can continue to run.
Explorations in Other Businesses
It is becoming more common to see collective ownership in other types of businesses as well. Participants provide physical resources, and in return they get incentives like token rewards or discounts to some of the services the network provides.
Some AI models are starting to catch on to the benefits of decentralized computing. DeepSeek, for example, has begun to leverage the decentralized model for its open-source computing. This allows it to source that computation out and reduce overhead.
In some respects, DePIN is similar to the cloud, which also has decentralized components. But there is a key difference in that cloud systems are usually tied to a centralized infrastructure in single, large data centers.
Next Steps
Hugentobler recommends that financial institutions and other technology providers begin looking at decentralized computing networks for tasks such as AI model training, high-volume data processing, and real-time financial computations. Just as bitcoin mining matured from central processing units to highly efficient application-specific integrated circuits, decentralized computing is likely to follow a similar trajectory, driving greater performance and cost efficiency over time.
Decentralized computing has matured to where organizations should already have it on their road map for further study and understanding. But only through hands-on experimentation will companies be able to tweak and enhance decentralized systems, determine which blockchains to use, manage incentives, and hash out other issues that need to be considered in launching a pilot. It’s a lengthy process, but all indications are that computation-heavy organizations will greatly benefit from the process.
