io.net Launches Beta Network to Harness Global GPU Power for AI Computing
Summary:
Start-up io.net has launched a beta for a decentralized physical infrastructure network (DePIN), set to incorporate over 100,000 GPUs from data centers and private clusters. The platform offers a novel solution to harness GPU computing power from diverse sources for advancing AI and machine learning. The company has collaborated with Render, enabling access to a vast range of computing resources. Primarily aimed at machine learning engineers and businesses, this innovation allows users to customize their GPU requirements to enhance efficiency and meet specific needs.
A new beta platform for a decentralized physical infrastructure network (DePIN) is set to incorporate over 100,000 GPUs from various data centers and private clusters, as revealed by start-up io.net. The pioneering tech company has created a unique network, which collects GPU computing power from global data centers, cryptocurrency miners, and decentralized storage providers to enhance machine learning and AI computing capabilities. The beta platform's launch was announced at the Amsterdam Solana Breakpoint conference in tandem with a new partnership with Render Network.
io.net's COO, Tory Green, alongside Angela Yi, the head of business development, gave a detailed explanation to Cointelegraph about what sets io.net’s DePIN apart from the pack in the cloud and GPU computing market. Green demarcates organizations like AWS and Azure as entities that possess their own GPU supplies and lease them. On the other hand, peer-to-peer GPU aggregators were developed to address GPU shortages, yet as Green explained, they promptly faced similar issues.
Green holds that existing infrastructure providers in the wider Web2 industry are attempting to tap into GPU computing from unused sources. However, none of them cluster GPUs the way Ahmad Shadid, the founder of io.net has innovated. Green comments that these providers typically work as a single instance and rarely cluster. Oftentimes, if there's a clustering option listed on provider's websites, a salesperson has to manually find out what's available from different data centers.
In contrast, Web3 firms like Render, Filecoin and Storj provide decentralized services that focus on something other than machine learning, and this is where io.net’s offerings can make a substantial contribution to the Web3 sphere. According to Green, close competitors in terms of functionality are AI-oriented solutions like Akash network, which clusters between 8 to 32 GPUs, and GenSyn, which is developing its own machine learning compute protocol to provide a peer-to-peer “supercluster” of resources.
Green pitched io.net’s solution as one-of-a-kind, with the capacity to cluster across diverse geographic locations within a matter of minutes. This was demonstrated by Yi, who successfully formed a GPU cluster from different networks and locations for a live demo.
In terms of their use of the Solana blockchain for processing payments to GPU computing providers, both Green and Yi agreed that no other network could handle the volume of transactions and inferences that io.net will facilitate. Their partnership with Render, a well-established DePIN network of distributed GPU suppliers, will offer io.net access to a vast range of computing resources already deployed on Render's platform. Aimed at sourcing GPU rendering computing faster and more economically than centralized cloud solutions, Render's network will benefit from io.net’s clustering capabilities.
In order to incentivize GPU resource providers, io.net are initiating a $700,000 program. This will also allow Render's nodes to extend their GPU capacity beyond graphical rendering to AI and machine learning applications. The target audience of this program are those with consumer-grade GPUs, specifically hardware from Nvidia RTX 4090s and below. Yi emphasized that many global data centers are only utilizing between 12% to 18% of their GPU capacity due to inefficient markets; these underutilized data centers are a focus for the io.net team.
The primary targets of io.net’s infrastructure are machine learning engineers and businesses, offering an extensively modular user interface that tailors GPU requirements, location, security parameters, and other essential metrics to user needs.
Published At
11/7/2023 1:59:12 PM
Disclaimer: Algoine does not endorse any content or product on this page. Readers should conduct their own research before taking any actions related to the asset, company, or any information in this article and assume full responsibility for their decisions. This article should not be considered as investment advice. Our news is prepared with AI support.
Do you suspect this content may be misleading, incomplete, or inappropriate in any way, requiring modification or removal?
We appreciate your report.