Key Points
- •Distributed networks of GPUs coordinated by crypto token incentives rather than corporate ownership
- •Aethir generated $128 million in revenue in 2025 across 440,000 GPU containers in 94 countries
- •io.net aggregates 327,000 verified GPUs at up to 70% cost savings versus centralized cloud providers
- •The DePIN sector spans 250 projects with $19-30 billion in combined market cap
- •Solves the GPU bottleneck by unlocking idle hardware worldwide through economic incentives
The GPU Bottleneck
The intelligence explosion runs on compute. Every model trained, every inference served, every agent deployed requires GPU cycles. Demand for AI compute is doubling faster than centralized providers can build data centers, and building a data center takes years. The result is a global GPU shortage that constrains the pace of AI progress.
Cloud providers like AWS, Azure, and Google Cloud control the majority of available GPU capacity, but even they cannot keep up. Waitlists for high-end NVIDIA hardware stretch months. Enterprises pay premium prices for capacity that may not be available when they need it. The centralized supply model has a structural scaling problem: it requires massive upfront capital, long construction timelines, and geographic concentration.
The Decentralized Alternative
Decentralized compute networks solve this by aggregating GPU capacity from independent providers worldwide, coordinated by crypto token incentives rather than corporate procurement.
The mechanism is straightforward. GPU owners contribute hardware to the network. Smart contracts handle job scheduling, payment, and result verification automatically. Contributors earn tokens proportional to the compute they provide. Customers access a global pool of GPUs at market-determined prices without enterprise sales cycles, procurement approvals, or geographic constraints.
This creates a compute market that scales at the speed of token incentives rather than the speed of construction.
The Networks
Aethir is the revenue leader. It generated $128 million in revenue in 2025 by coordinating 440,000 GPU containers across 94 countries for 150 enterprise clients. Revenue grew 22% quarter over quarter in Q3 alone, with over 1.5 billion compute hours delivered.
io.net has aggregated 327,000 verified GPUs across 130 countries, offering up to 70% cost savings versus centralized providers. It generated over $20 million in verifiable on-chain revenue in 2025.
Akash Network grew deployments 466% to 3.1 million in 2025, with GPU utilization trending above 80%. Its Starcluster initiative combines protocol-owned compute with the decentralized marketplace.
Render Network operates 5,600 GPU nodes at 85-95% utilization and launched Dispersed.com in December 2025 specifically for AI inference, with 600 open-weight models available.
Gensyn extends the decentralized model to training, with a public testnet that reduces distributed training time by 55% through its SkipPipe protocol.
DePIN: The Broader Category
Decentralized compute is part of a broader category called DePIN, Decentralized Physical Infrastructure Networks. DePIN projects use token incentives to build real-world infrastructure: compute, storage, wireless coverage, energy, mapping.
The DePIN sector spans 250 projects with $19-30 billion in combined market cap and over 10 million devices deployed across 199 countries. AI-related DePIN projects represent 48% of the sector by market cap, reflecting the enormous demand for compute infrastructure.
Why Decentralization Matters for AI
Centralized compute creates concentration risk. A small number of cloud providers control the infrastructure that the entire AI industry depends on. They choose who gets access, at what price, under what terms. This creates dependencies that become dangerous as AI becomes more consequential.
Decentralized compute distributes this power. No single entity can throttle access to the intelligence explosion. A model developer cut off by one provider can access the same GPUs through a permissionless market. Geographic decentralization provides resilience against regulatory restrictions, natural disasters, and infrastructure failures.
Decentralized Compute and the Singularity
The intelligence explosion requires compute that scales faster than any single organization can build. Token-incentivized networks unlock existing GPU capacity worldwide, converting idle hardware into productive infrastructure through economic incentives. The result is a compute substrate that grows with demand, operates without gatekeepers, and makes no distinction between researchers at a frontier lab and an independent developer working from a laptop. This is how the intelligence explosion gets its fuel.
