io.net:分布式 GPU 聚合io.net: aggregated distributed GPUs
交集:把独立数据中心、矿业与私有集群的 GPU 聚合起来,按需供给训练/推理;价格与可用性随供给波动。
适用范围:官方称可按需提供 H100/A100 与整簇、分钟级部署、全球分布式网络与可选机密计算;不承诺固定库存或单一区域。
未解决:跨供给方的网络带宽与延迟、数据安全与合规、任务迁移与故障恢复、以及「70% cheaper」这类对比的基准条件需自行核验。Intersection: aggregate GPUs from independent data centers, mining operations and private clusters, then supply them on demand for training/inference; price and availability vary with supply.
Scope: the official site advertises on-demand H100/A100 and full clusters, minute-level deployment, a globally distributed network and optional confidential compute; it does not promise fixed inventory or a single region.
Open: cross-provider bandwidth and latency, data security/compliance, job migration and failure recovery, and the basis of "70% cheaper" comparisons need independent verification.
关键参数Key specs
- 协议状态 / status
- 已部署
- 供给 / supply
- 独立数据中心 + 矿业 + 私有集群
- 卡型 / GPUs
- H100 / A100 / 多类
- 机密计算 / confidential
- 可选
官方原文怎么说What the source says
H100s, A100s, and full clusters on demand.
Securely connect to GPUs from independent data centers, mining operations, and private clusters worldwide.
globally distributed network, and available confidential compute
资料与许可Source and licence
- 资料时间Source checked
- 2026-11-19
- 许可 / 类型Licence / kind
- official-specification
- 许可说明Licence evidence
- https://io.net/
- 类目Category
- 去中心化推理Decentralised inference