Observed arrival · 2026-09-03
HoloNear, the blind supercomputer
HoloNear proposes a privacy-preserving compute network that runs model training and inference across idle machines.
Field notes
The homepage lays out both sides of the proposed network: developers submit inference or training jobs through an OpenAI-compatible base URL, while providers run a node with a terminal command or app. Its privacy model depends on hardware-attested enclaves, sealed tensors, and hardware-bound keys rather than trust in the host machine. The page also publishes preliminary rates for several named models and training workloads, while distinguishing public-rate earning ceilings from reported payouts.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
One card from the complete issue