top of page


DePIN: An inclusive approach for building physical and digital infrastructure

Decentralized Physical Networks (DePIN) typically employ blockchain technology to coordinate and manage the deployment and operation of real-world physical infrastructure and hardware networks. The key aspect is that they harness a vast number of individuals and crypto-economic incentives to fund the underlying infrastructure. Unlike centralized entities that follow a top-down approach when deploying hardware infrastructure, crypto-economic protocols can align millions of individuals not only to deploy and operate infrastructure in a trustless and permissionless manner but also to enhance efficiency, security, and resilience. This creates a flywheel effect: token rewards incentivize supply-side participants, attracting new end-users as the network grows, generating fees as more end-users utilize the network and services, and increasing the value of the token, thereby incentivizing more users on the hardware supply side. 


Some may wonder why crypto incentives are necessary. Firstly, permissionless and borderless protocols can expand their network globally across various legal jurisdictions, enabling rapid scale. Furthermore, a network or system controlled and owned by distinct parties fosters loyalty and aligns incentives, bringing collective ownership rights to the table. Lastly, by using tokens and cryptocurrencies, these systems enable P2P micropayments and cross-border payments almost instantaneously and at minimal cost. 


You might argue that token incentives alone do not sustain a healthy business model, and you would be correct. While, in the short term, they may address the so-called "cold start problem" to bootstrap a capital-intensive network, in the long run, most of these community-built systems benefit from significantly lower operating expenses (OPEX) and capital expenses (CAPEX) compared to their centralized counterparts. 


Take Helium, for instance, the most prominent protocol providing an IoT and 5G network with approximately 380,000 hotspots online, deployed worldwide — a network bootstrapped by the community. Another noteworthy protocol is Filecoin, a decentralized, peer-to-peer storage network built on top of IPFS. We will delve into the protocol and its ecosystem in more detail in the next episode, as we see numerous use cases for building applications and business models on top of it. 


Categorically, for DePIN projects, they have successfully created robust supply-side incentives to bootstrap a capital-intensive network in a zero-interest environment. However, the demand side of the network often fell short, with insufficient customers willing to pay for the offered services. 


For a comprehensive overview of the DePIN landscape, we recommend referring to the DePIN sector map by Messari or the newly proposed classification framework of DePIN and DeREN by Variant Fund


Key concepts that pique our interest in this sector: 


Regarding physical infrastructure networks, we specifically support protocols that collect valuable data and/or provide services by leveraging affordable or existing hardware, such as sensors in smartphones. In a high-interest environment, it becomes much more challenging to bootstrap capital-intensive hardware networks and to avoid value outflow due to the need to compensate for CAPEX or OPEX. Furthermore, making use of non-static providers can be a key strategy to cover larger areas and reduce reliance on vast amounts of capital-intensive hardware. Such protocols should aim to keep user participation requirements as low as possible to onboard many providers quickly and scale the network rapidly. 


On the other hand, (digital) resource networks (e.g., storage and compute networks) mainly compete with existing centralized and web2 incumbents. Since existing resource networks primarily compete on lower price levels by subsidizing their offerings or services, this approach is not sustainable in the long run. They must tap into markets that these incumbents are not currently serving. 


Let's take compute networks as an example, as the AI hype has made GPUs one of the scarcest resources in the digital world. Accessing computing power can be a significant bottleneck in training your model, and decentralized compute networks could fill this gap by incentivizing the contribution of idle GPU and CPU capacities to the network. However, this alone is insufficient. These networks must also focus on addressing other challenges faced by modern ML models. One of the challenges modern ML models face is that the end user tends to have trust issues and fears misuse of others. The lack of transparency and provability is a significant issue, as today's ML models are black boxes to users. Therefore, we must create a trust framework that makes machine learning trustworthy and reliable with on-chain verifications. Scalability and privacy issues that arise with this step can be resolved by incorporating cryptographic technologies such as zero-knowledge (zk), MPC, or FHE. 

The reading above provides just a taste of what currently excites us the most. But that doesn't mean we're not open to other sectors and venture ideas. If you're building something in this particular sector that you are extremely passionate about, don't hesitate to apply. We can’t wait to hear from you! 

bottom of page