What people build on CE

Serious systems, on shared compute.

CE is a substrate for real work — research, production, content distribution, and infrastructure that has to keep running when everything else goes down. The games below are the stress tests: if the mesh can hold a physically-simulated universe with thousands of players, it can hold your workload.

Flagship demos

Built to break the mesh — and prove it doesn't.

Spacegame

An infinite, procedurally generated universe where every ship is physically simulated in 2D. Interest-managed across the mesh so thousands can share one galaxy.

live · open →

CE Cast

A live multicast studio: send one stream from any device — even a slow laptop — and an encoder node does the heavy lifting, fanning out to YouTube, Twitch, Kick, and X at once.

live · open →

Cerena

A first-person procedural mage RPG built to hold thousands of concurrent players in one authoritative world — make your own spells, parkour, a hot-reloadable tech tree, and proximity-replicated state. The hardest realtime test of mesh placement, latency, and fault tolerance.

in development
The point

Compute as a commons, for work that matters.

Every category below composes the same primitives — identity, mesh, ledger, capabilities, compute — with no node forks. Your app is one bundle; it joins the mesh like every other.

Research

Protein folding, climate ensembles, genomics, astronomy, ML training — pooled compute pointed at the hardest problems.

AlphaFold · GROMACS · TESS

Production systems

CI runners, render farms, batch inference, headless browsers — sandboxed jobs billed by the second on compute you don't own.

Docker · gVisor · WASM

Content distribution

Content-addressed storage and a mesh CDN — publish once, served from wherever it's closest, verifiable by hash.

ce-storage · ce-cdn

Social & collaboration

Chat, kanban, notes, drives — local-first and end-to-end encrypted, converging across devices over the mesh.

ce-chat · ce-board · ce-notes

Wartime & resilience aid

Infrastructure that survives blackouts and blockades: offline-first comms and content distribution with no central server to seize or sever.

mesh-first · no chokepoint

On-prem & private AI

Distributed LLM inference where data never leaves the LAN — PHI, classified, or proprietary work stays put.

ce-infer · ce-exo