- The Partnership enables affordable, secure, on-demand GPU clusters for AI workloads.
- Phala integrates TEE technology using Nvidia GPUs for advanced AI capabilities.
- IO Cloud offers on-demand GPU clusters, democratizing access and reducing costs.
GPU cloud network io.net has entered into a strategic partnership with Phala Network to advance secure computation and decentralized artificial intelligence (AI). This collaboration will enable Phala to expand its decentralized AI capabilities by leveraging GPU hardware through io.net’s advanced cloud network, IO Cloud.
As part of the partnership, Phala Network will integrate its Trusted Execution Environment (TEE) nodes with io.net’s infrastructure, enabling secure and computationally intensive decentralized applications (dapps) to thrive.
Integrating Phala’s TEE nodes with io.net’s infrastructure
Phala Network, which uses Nvidia H100 and H200 GPUs to power its TEE technology, will benefit from enhanced access to these critical hardware resources, strengthening the foundation of its decentralized AI ecosystem.
In a notable achievement, Phala recently set a benchmark for TEE-enabled GPUs, raising the standard for decentralized AI processing.
By extending TEE capabilities to AI accelerators like GPUs, this collaboration ensures that sensitive AI workloads are handled with strong cryptographic protection.
Nvidia’s H100 Tensor Core GPUs, featuring confidential computing technologies such as encrypted memory and secure boot, are pivotal in Phala’s security strategy. The integration with io.net’s cloud infrastructure will provide Phala with the necessary computational power to support the training and execution of large-scale AI models, including LLaMA 3 and Microsoft Phi.
Since its mainnet launch in 2021, Phala has built an infrastructure of over 40,000 TEE CPU nodes, empowering web3 developers to offload complex computations securely. These nodes are essential in maintaining data privacy, offering verifiable proofs, and enabling a range of web3 applications, from AI-driven agents to decentralized social apps.
Deploying decentralized on-demand GPU clusters
With io.net’s IO Cloud, users can deploy decentralized GPU clusters on demand, eliminating the need for costly hardware investments. The platform offers up to 90% cost savings compared to traditional cloud providers, democratizing access to advanced computational resources.
The partnership will also focus on research and benchmarking, starting with cutting-edge Nvidia H100 and H200 GPUs, as both companies explore deploying autonomous AI agents and integrating Phala Network’s TEE technology into io.net’s decentralized GPU network.