Solana-based decentralized infrastructure provider io.net, known for allowing users to rent out GPU power for financial gain, has made a significant change in leadership just two days before the launch of its token.
Ahmad Shadid, a co-founder of io.net, made the decision to step down “effective immediately” and has been replaced by another co-founder and former chief operating officer, Tory Green.
The Solana-based AI project aggregates GPU supply to establish a network that enables machine learning startups to access computing power at a fraction of the cost compared to traditional cloud services.
In a statement posted on X on June 9, Shadid mentioned that he was stepping down to eliminate distractions and allow io.net to focus on growth and success, without addressing the specific “allegations” that had been circulating. Critics speculated that Shadid had misrepresented the number of GPU chips available on io.net.
The network faced a setback on April 28 when it experienced a GPU metadata attack, causing the number of active GPU connections to temporarily drop from 600,000 to 10,000.
The launch of the io.net token, IO, is scheduled for June 11 on Binance’s Launchpool at 12:00 am UTC. A total of 95,000,000 IO tokens will be released initially, with a maximum supply of 800,000,000 IO tokens possible.
Concerns arose about Shadid potentially selling off his IO coins at launch and disappearing, but he clarified that his tokens are locked up for four years and cannot be sold until June 2025. He also pledged to contribute one million IO tokens to the Internet of GPUs Foundation to support the ecosystem’s growth.
It remains unclear if Shadid will maintain any involvement with the io.net ecosystem, as the company has not provided a response to inquiries from Cointelegraph. Green, the new CEO, hinted at additional leadership changes in the near future.
Green emphasized that the token launch signifies a new phase of growth for the network, with a continued focus on building the largest decentralized AI compute network. Io.net currently has approximately 20,000 cluster-ready GPUs and caters to various AI-focused companies for AI inference and model training workloads.