IO.NET

• io.net aggregates underutilised GPUs into clusters that machine learning teams can rent on demand.

• It draws capacity from independent data centres, crypto miners and individual operators.

• IO is used for payments to suppliers and for staking by network participants.

• Fixed maximum supply with emissions to suppliers that decline over time.

• Suppliers stake IO as collateral against reliability commitments.

Pros:

• Aggregating existing idle capacity avoids the capital cost of building data centres.

• Cluster deployment is faster than procurement from conventional cloud providers.

• Built on Solana, keeping settlement costs low for frequent small payments.

Cons:

• Shortly after launch the network faced reporting that a substantial share of listed GPUs were duplicated or non-functional, and it introduced stricter verification in response.

• Consumer and mining GPUs are poorly suited to large-scale model training, which needs tightly interconnected hardware.

• Competes against hyperscale providers with far better networking.

• Exposure to decentralised GPU aggregation at meaningful scale.

• Cost advantage is real for workloads that tolerate loose interconnection.

• The supply verification episode is important context on how capacity figures should be read.

• Clusters assembled from geographically distributed independent suppliers, deployable in minutes rather than through a procurement cycle.

• GPU capacity is scarce and concentrated among a few cloud providers with long waitlists.

• Aggregating idle hardware creates supply outside that bottleneck, at the cost of weaker interconnection.

• Founded by Ahmad Shadid, with leadership changes since launch.

• Backed by Hack VC, Multicoin Capital and Solana Ventures.

• No public ICO. IO was distributed via airdrop and exchange programmes in June 2024.

• Capacity figures require independent verification given the post-launch reporting issues.

• Distributed consumer GPUs cannot serve large training workloads well.

• GPU scarcity easing would remove the primary demand driver.

• Leadership turnover since launch.