The network
Gennode is a private, torrent-like node network for heavy bio & health computation — a decentralized layer focused on one vertical. Nodes contribute compute and storage; the goal is to analyze sensitive biological data privately and at scale.
Three layers
- Compute — nodes run jobs: DNA & genome analysis, protein folding, drug discovery, disease research and multi-omics. → Use cases
- Data (torrent-like) — planned — on the roadmap, datasets, model weights and results will be encrypted, chunked, content-addressed and served peer-to-peer across nodes. → Data layer
- Models — bio AI models the network runs to analyze biological data (protein folding, genomics) — open models today; a purpose-built Gennode bio model served across nodes over time.
How a job flows (planned)
The full flow below describes the network once real compute and the data layer are live:
Demand (researcher / lab / app / a job request)
→ Coordinator (matches job + data → capable nodes, verifies the result)
→ Nodes (pull encrypted data from the swarm, compute, return result)
→ Result stored in the data layer + reward in $GENNODEPrivacy (planned — a core design goal for bio data)
These privacy properties are part of the roadmap, not yet live:
- Data is intended to be encrypted client-side before it ever leaves the owner, so nodes store encrypted shards they cannot read.
- Compute roadmap: confidential computing (GPU TEE) + sealed containers so operators cannot see data even while it is being processed.
- Owners control access with keys; nothing is sold or profiled.
Anti-sybil
Today, fairness comes from wallet linking, a CPU benchmark that proves a node runs on real hardware, and a device fingerprint (one device = one node = one wallet). Storage proofs and cluster analysis are planned.
Node roles & rewards
Today, a node earns $GENNODE points for uptime and proving real hardware. As the network matures, nodes will also earn for compute (running jobs) and storage (holding & serving data shards, proven) — see the Roadmap.