- GPU-as-a-service provider, Aethir, is leveraging the Ethereum network to offer its product to a variety of potential users.
- The company plans to aid artificial intelligence developers by allowing them to use idle GPU resources from participants.
- This is a form of decentralised physical infrastructure (DePIN), which is playing a major role in blockchain’s evolution into the mainstream.
Ethereum has added another major partnership to its list, with decentralised physical infrastructure (DePIN) service Aethir joining its ranks. The collaboration was announced earlier this week, with Aethir planning to launch its product on the Ether Mainnet – an interesting move, considering that much of the community is enamoured with L2 Ether protocols.
Nevertheless, the introduction of Aethir’s cloud computation software to the DePIN world is set to spark discussion over how vital this technology can be for the future of Web3.
Related: Blockchain Could Boost Australian Economy by $60 Billion, Reverse Productivity Slump, Says MP Andrew Charlton
Aethir Set to Improve Efficiency for AI Developers
DePINs are growing in importance among the blockchain world. They intend to replace traditional, centralised infrastructures (such as databases) by using a peer-to-peer network – much how a decentralised transaction works.
DePINs also incentivise community engagement, rewarding participants for keeping the network secure. For now, most DePINs are focussed on decentralising data storage, an industry where the prominent services are owned by a select few companies (Google, Amazon, Microsoft).
Aethir’s co-founder, Mark Rydon, discussed the importance of DePINs in disrupting current infrastructure solutions:
DePIN’s architecture itself has good high availability and fault-tolerant basic capabilities. Distributed computing power avoids the problem of service nodes being too centralized, resulting in insufficient single-point service capability and low fault tolerance rate.
Mark RydonAethir is specifically targeting GPU-as-a-service, offering AI developers the ability to “rent” the computational power of other people’s GPUs. In return for lending their GPUs, participants will be compensated in a native cryptocurrency.
By providing a scalable framework for redistributing idle compute resources, we can empower more innovation in the rapidly evolving domains of AI, ML and cloud gaming.
Mark RydonAethir’s Testnet comprised over 500,000 users and included support from Nvidia, Super Micro, HPE and Foxconn to ensure the network’s security.
Credit: Source link
