en
Back to the list

io.net partners with Phala Network to enhance secure computation and AI

source-logo  invezz.com 26 September 2024 13:37, UTC

io.net, a decentralized GPU cloud network, has entered into a strategic partnership with Phala Network to advance secure computation and decentralized AI.

The collaboration will leverage io.net’s cloud platform, IO Cloud, to provide Phala Network with access to high-performance GPU hardware, enhancing its ability to support TEE (Trusted Execution Environment) CPU nodes.

This partnership is a significant step toward enabling more advanced and secure AI computations within decentralized ecosystems.

Expanding Phala’s decentralized AI capabilities with GPU access

Phala Network, which relies on Nvidia H100 and H200 GPUs to power its TEE infrastructure, will benefit from io.net’s GPU cloud service to enhance its AI computing capacity.

By integrating with io.net’s IO Cloud, Phala will gain access to the necessary GPUs to process computationally intensive decentralized applications (apps).

This access is vital for running large-scale AI models like LLaMA 3 and Microsoft Phi, ensuring that Phala Network can support cutting-edge decentralized AI systems securely and efficiently.

In August, Phala Network set a new benchmark for TEE-enabled GPUs, marking a major advancement in decentralized AI.

By expanding the capabilities of TEEs to include AI accelerators like GPUs, Phala’s collaboration with io.net will enhance its infrastructure to handle sensitive AI workloads with advanced cryptographic protections.

This will further Phala’s mission of building a secure, privacy-focused decentralized network for AI-driven agents and web3 applications.

Nvidia H100 GPUs to power secure AI applications

A key element of this collaboration is the utilization of Nvidia’s H100 and H200 Tensor Core GPUs.

These GPUs, equipped with confidential computing features such as encrypted memory and secure boot, will form the backbone of Phala Network’s secure AI infrastructure.

Access to these high-performance GPUs via io.net’s IO Cloud will allow Phala to train and deploy large AI models with robust security features.

Since its mainnet launch in 2021, Phala Network has established an infrastructure of over 40,000 TEE CPU nodes, providing web3 developers with secure, off-chain computation capabilities.

These nodes have been instrumental in maintaining privacy and security across a variety of decentralized applications, from social platforms to AI-driven systems.

The new partnership with io.net will ensure that Phala continues to scale its AI capabilities while maintaining its focus on security.

Research and benchmarking to push AI technology forward

The collaboration between io.net and Phala Network will focus on joint research, testing, and benchmarking.

By starting with the latest Nvidia H100 and H200 GPUs, the partners aim to push the boundaries of decentralized AI.

Research will include deploying Phala Network’s autonomous AI agents and integrating its TEE hardware into io.net’s GPU cloud network.

This joint effort aims to provide a scalable, secure solution for decentralized AI applications.

IO Cloud, io.net’s decentralized GPU platform, offers users access to powerful GPU clusters on demand, providing significant cost savings compared to traditional cloud services.

By democratizing access to advanced computing power, IO Cloud enables AI developers to build, test, and deploy AI models without the need for expensive hardware or complex infrastructure.

The post io.net partners with Phala Network to enhance secure computation and AI appeared first on Invezz

invezz.com