Where the network’s machines are
HyperCluster’s main server in Singapore keeps a copy of every file; Quipu machines and members’ PCs keep files and run AI with it. Our blockchain nodes, the squares in other cities, keep the chain running with it. Only machines online now are shown. Each circle is a city: the bigger it is, the more machines are near it, and its colours show what they share.
9 machines online now, with HyperCluster’s own servers. The boxes add up what they share with the network now; TOPS is an AI chip’s rated speed. Blockchain nodes are HyperCluster’s own servers that keep the chain (the squares on the map); the validators among them sign its blocks. See it on the explorer.
To keep owners private, each machine is shown only at the city of the Cloudflare data centre it connects through (usually the nearest one), never at its address and never with its owner’s name. HyperCluster’s own servers are shown where they are: the main server in Singapore, and our blockchain nodes. Counted at 01:24 UTC, again every minute.
Cities
- HyperCluster’s main serverSingapore · keeps a copy of every file; keeps a copy of the chain1 storage0 AI chip0 graphics card1 blockchain node
- HyperCluster’s blockchain nodeHong Kong · a validator, signing the chain’s blocks0 storage0 AI chip0 graphics card1 blockchain node
- HyperCluster’s blockchain nodeSingapore · a validator, signing the chain’s blocks0 storage0 AI chip0 graphics card1 blockchain node
- HyperCluster’s blockchain nodeSan Jose, United States · a validator, signing the chain’s blocks0 storage0 AI chip0 graphics card1 blockchain node
- HyperCluster’s blockchain nodeNuremberg, Germany · a validator, signing the chain’s blocks0 storage0 AI chip0 graphics card1 blockchain node
- Kuala Lumpur, Malaysia4 machines online now4 storage1 AI chip2 graphics card0 blockchain node
How the map works
Every machine connects to HyperCluster through Cloudflare, a service with data centres in about 300 cities. A machine is shown at the city of the data centre its connection arrives at, which is usually the one nearest to it.
A machine is counted while it is online and shares at least one of these with the network: disk space where it keeps copies of files, an AI chip (NPU) that has been checked and approved for work, or a graphics card (GPU) its owner shares that passed the network’s test. A machine that is offline is not shown or counted.
HyperCluster’s main server, in a data centre in Singapore, keeps a copy of every file the network stores. It is shown where it is, as a storage node, while its storage is running.
HyperCluster’s blockchain nodes keep a copy of our chain, and the validators among them sign its blocks: the main server and our servers in other cities. They are ours, so each is shown where it is, as a square, while it is running.