The Intersection of AI and DePIN: The Rise of Decentralized GPU Networks

Since 2023, AI and DePIN have shown a thriving trend in the Web3 field. The market value of AI has reached 30 billion dollars, while the market value of DePIN is approximately 23 billion dollars. These two fields encompass numerous different protocols that serve various needs. This article will explore the intersection of the two and study the development of protocols in this area.

In the AI technology stack, the DePIN network provides practicality for AI through computing resources. The development of large tech companies has led to a shortage of GPUs, making it difficult for other developers building AI models to obtain enough GPUs for computation. This often forces developers to choose centralized cloud providers, but the need to sign inflexible long-term high-performance hardware contracts results in inefficiency.

DePIN essentially offers a more flexible and cost-effective alternative that uses token rewards to incentivize resource contributions that align with network objectives. In artificial intelligence, DePIN crowdsources GPU resources from individual owners to data centers, forming a unified supply for users who need access to hardware. These DePIN networks not only provide customization and on-demand access for developers requiring computing power but also offer additional income for GPU owners.

There are numerous AI DePIN networks on the market, and this article will explore the role and objectives of each protocol, as well as some specific highlights they have already achieved.

AI DePIN Network Overview

Render is a pioneer of a P2P network that provides GPU computing power, previously focused on rendering graphics for content creation, and later expanded its scope to include computational tasks from neural radiance fields (NeRF) to generative AI by integrating tools like Stable Diffusion.

Features of Render:

1. Founded by the cloud graphics company OTOY with Oscar-winning technology.
2. GPU networks have been utilized by major entertainment companies such as Paramount Pictures, PUBG, and Star Trek.
3. Collaborate with Stability AI and Endeavor to integrate AI models with 3D content rendering workflows using Render's GPU.
4. Approve multiple computing clients and integrate more GPUs from DePIN networks.

Akash positions itself as a "super cloud" alternative to traditional platforms that support storage, GPU, and CPU computing. With developer-friendly tools like the Akash container platform and Kubernetes-managed compute nodes, it enables seamless deployment of software across environments, making it capable of running any cloud-native application.

Features of Akash:

1. For a wide range of computing tasks from general computing to web hosting.
2. AkashML allows its GPU network to run over 15,000 models on Hugging Face, while integrating with Hugging Face.
3. Akash hosts some noteworthy applications, such as the LLM model chatbot from Mistral AI, the SDXL text-to-image model from Stability AI, and the new foundational model AT-1 from Thumper AI.
4. The platform for building the metaverse, deploying artificial intelligence, and federated learning is leveraging Supercloud.

io.net provides access to distributed GPU cloud clusters that are specifically designed for AI and ML use cases. It aggregates GPUs from data centers, crypto miners, and other decentralized networks.

Features of io.net:

1. Its IO-SDK is compatible with frameworks such as PyTorch and Tensorflow, and its multi-layer architecture can automatically and dynamically scale according to computational demands.
2. Supports the creation of 3 different types of clusters, which can be started within 2 minutes.
3. Strong collaborative efforts to integrate the GPUs of other DePIN networks, including Render, Filecoin, Aethir, and Exabits.

Gensyn provides GPU computing power focused on machine learning and deep learning computations. It claims to achieve a more efficient verification mechanism by combining concepts such as proof of work for validating jobs, graph-based precise location protocols for re-running validation jobs, and Truebit-style incentive games involving staking and slashing of computing providers.

Features of Gensyn:

1. The expected hourly cost of a V100 equivalent GPU is approximately $0.40/hour, resulting in significant cost savings.
2. Through proof stacking, the pre-trained base model can be fine-tuned to complete more specific tasks.
3. These foundational models will be decentralized, globally owned, and will provide additional functionality beyond the hardware computing network.

Aethir is specifically equipped with enterprise GPUs, focusing on compute-intensive areas, mainly artificial intelligence, machine learning (ML), cloud gaming, and more. The containers in its network act as virtual endpoints for executing cloud-based applications, transferring workloads from local devices to the containers to achieve low-latency experiences. To ensure high-quality service for users, they move the GPUs closer to the data sources based on demand and location, thereby adjusting resources.

Features of Aethir:

1. In addition to artificial intelligence and cloud gaming, Aethir has also expanded into cloud phone services and collaborated with APhone to launch a decentralized cloud smartphone.
2. Established extensive partnerships with large enterprises such as NVIDIA, Super Micro, HPE, Foxconn, and Well Link.
3. Multiple partners in Web3, such as CARV, Magic Eden, Sequence, Impossible Finance, etc.

Phala Network serves as the execution layer for Web3 AI solutions. Its blockchain is a trustless cloud computing solution designed to address privacy issues by using its trusted execution environment (TEE). Its execution layer is not used as the computing layer for AI models, but instead enables AI agents to be controlled by smart contracts on the chain.

Features of Phala Network:

1. Serve as a verifiable computing coprocessor protocol, while also enabling AI agents to access on-chain resources.
2. Its AI agent contracts can access top large language models such as OpenAI, Llama, Claude, and Hugging Face through Redpill.
3. The future will include zk-proofs, multi-party computation (MPC), fully homomorphic encryption (FHE), and other multiple proof systems.
4. Future support for H100 and other TEE GPUs to enhance computing power.

Project Comparison

| | Render | Akash | io.net | Gensyn | Aethir | Phala | |--------|-------------|------------------|---------------------|---------|---------------|----------| | Hardware | GPU & CPU | GPU & CPU | GPU & CPU | GPU | GPU | CPU | | Business Focus | Graphics Rendering and AI | Cloud Computing, Rendering and AI | AI | AI | Artificial Intelligence, Cloud Gaming and Telecommunications | On-chain AI Execution | | AI Task Type | Inference | Training and Inference | Training and Inference | Training | Training | Execution | | Work Pricing | Performance-Based Pricing | Reverse Auction | Market Pricing | Market Pricing | Bidding System | Equity Calculation | | Blockchain | Solana | Cosmos | Solana | Gensyn | Arbitrum | Polkadot | | Data Privacy | Encryption& Hashing | mTLS Authentication | Data Encryption | Secure Mapping | Encryption | TEE | | Work Fees | 0.5-5% per job | 20% USDC, 4% AKT | 2% USDC, 0.25% reserve fee | Low fees | 20% per session | Proportional to the staked amount | | Security | Render Proof | Proof of Stake | Proof of Computation | Proof of Stake | Render Capability Proof | Inherited from Relay Chain | | Proof of Completion | - | - | Time-Lock Proof | Proof of Learning | Render Work Proof | TEE Proof | | Quality Assurance | Dispute | - | - | Verifier and Reporter | Checker Node | Remote Proof | | GPU Cluster | No | Yes | Yes | Yes | Yes | No |

The availability of clustering and parallel computing is crucial for training complex AI models. Most projects have now integrated clusters for parallel computing. io.net has collaborated with other projects such as Render, Filecoin, and Aethir to incorporate more GPUs into its network and has successfully deployed over 3,800 clusters in the first quarter of 2024.

Data privacy is a key issue in AI model development. Most projects use some form of data encryption to protect data privacy. io.net recently partnered with Mind Network to launch fully homomorphic encryption (FHE), allowing encrypted data to be processed without needing to decrypt it first. Phala Network introduced a Trusted Execution Environment (TEE), which can prevent external processes from accessing or modifying data.

The completion of proof and quality inspection is crucial for ensuring the quality of work. Both Gensyn and Aethir generate proofs to indicate that the work has been completed and perform quality checks on the completed computations. Render recommends using a dispute resolution process, where if the review committee finds issues with a node, that node will be penalized.

Hardware Statistics

| | Render | Akash | io.net | Gensyn | Aethir | Phala | |-------------|--------|-------|--------|------------|------------|--------| | Number of GPUs | 5600 | 384 | 38177 | - | 40000+ | - | | Number of CPUs | 114 | 14672 | 5433 | - | - | 30000+ | | H100/A100 Quantity | - | 157 | 2330 | - | 2000+ | - | | H100 Cost/Hour | - | $1.46 | $1.19 | - | - | - | | A100 cost/hour | - | $1.37 | $1.50 | $0.55 ( estimated ) | $0.33 ( estimated ) | - |

High-performance GPUs are crucial for AI model training. io.net and Aethir lead in the number of H100 and A100 GPUs, making them more suitable for large model computations. The cost of decentralized GPU networks is already much lower than centralized GPU services, opening up an oligopolistic situation for building more AI and ML use cases.

Consumer-grade GPUs/CPUs also play an important role in these networks. Projects like Render, Akash, and io.net can serve this segment of the market, providing developers with more options.

Conclusion

The AI DePIN field, while still relatively new, has shown strong development momentum. The increasing number of tasks and hardware executed in these decentralized GPU networks highlights the growing demand for alternatives to Web2 cloud provider hardware resources. Looking ahead, these decentralized GPU networks will play a key role in providing developers with cost-effective computing alternatives, making significant contributions to the future landscape of artificial intelligence and computing infrastructure.