- Published on
Why crypto's best infrastructure companies stopped looking like crypto?
- Authors
- Name
- DeepHorus
Storj, one of the few working service businesses in decentralized storage, filed for Chapter 11 bankruptcy in July. Its restructuring announcement called the underlying business 'strong and right-sized,' held back by 'legacy obligations from an earlier chapter,' and said it expects to keep operating without interruption. This is not an isolated case. At least three legacy DePIN companies have undergone restructuring, selling or splitting their businesses to decouple from the token component this summer. The old DePIN modelâincentivizing ordinary people to crowdsource hardware, bandwidth, and storageâis dead. The new playbook mirrors how every infrastructure sector has developed: build raw capacity ahead of demand, commoditize it, and let money flow to those selling guaranteed, contracted services on top.
Cloud computing ran this script perfectly. Amazon launched EC2 in 2006, selling raw compute by the hour. AWS now accounts for 18% of Amazon's revenue but 57% of its operating income. The services built on top are the ones accumulating value, not the infrastructure companies themselves. Netflix pays an estimated $1 billion annually for AWS compute and storage. Snowflake produces $4.5 billion in product revenue while owning zero data centers. Datadog earns $3.4 billion, Zoom $4.9 billion, and Airbnb $12.2 billionâall running on rented infrastructure their customers never see. Total public cloud spending hit $595.7 billion in 2024 and is projected to reach $723.4 billion in 2025, exceeding $1 trillion by 2027.
Telecom follows the same pattern. T-Mobile paid up to $1.35 billion for Mint Mobile, a phone brand that owns no towers, no spectrum, and no radiosâbut has Ryan Reynolds. Over 2,100 mobile virtual network operators (MVNOs) operate in about 100 countries, renting everything from infrastructure owners to offer services to end users. The tower companies (American Tower, Crown Castle, SBA) own the physical assets but earn modest margins compared to the service providers that leverage them.
The current AI boom complicates the curve. GPUs and power are scarce again, so profits are temporarily flowing back to the infrastructure layer. AWS's massive $200 billion CapEx for 2026, driven by AI, signals an infrastructure rebuild. But if history holds, this scarcity will resolve, and commoditization will continue. The absolute gap between infrastructure and services will keep widening.
For crypto, this means the winners won't look like crypto at all. The infrastructure companies that survive will be those that restructure like normal businesses, focusing on service revenue rather than token speculation. Storj's restructuring is a harbinger: the company expects to continue serving customers without interruption, detaching its operational health from its legacy obligations. The token component, once central to DePIN's value proposition, is becoming a liability.
Industry reactions are mixed. Some see the restructuring as a failure of the DePIN thesis. Others view it as natural maturation. 'This is the market correcting toward what actually works,' said one anonymous crypto infrastructure analyst. 'The businesses that provide real storage, bandwidth, or compute at competitive prices will survive. The token games are over.'
Looking ahead, the future of crypto infrastructure lies in becoming invisible. Just as AWS powers Netflix and Snowflake without consumers knowing, decentralized storage and compute will succeed when they integrate into traditional enterprise stacks. The companies that embrace thisâshedding their 'crypto' branding and focusing on contracted service agreementsâwill capture the value. The ones that cling to the old model will fade, leaving behind a graveyard of token projects. The playbook is written; crypto infrastructure is finally reading it.
