Mining

FIBRE Resurrected
Fast Block Relay

The Fast Internet Bitcoin Relay Engine is back. A patched Bitcoin Core client using UDP and Forward Error Correction to propagate blocks across the planet in tens of milliseconds — reducing centralization pressure on miners.

FIBRE node Block propagation

~30

ms avg propagation

The Problem: Latency Creates Centralization

Why block propagation speed matters for mining decentralization

▾

When a miner finds a block, every millisecond it takes to reach other miners matters. A miner who hears about a new block late is wasting hashrate mining on a stale tip. This creates orphan blocks, which represent lost revenue. The economics are simple: miners with better network connectivity have a structural advantage.

The centralization feedback loop: Large pools with more hashrate find more blocks. They hear about their own blocks instantly (zero latency to self). Smaller miners and solo operators, especially those far from major data centers, are systematically disadvantaged. This pushes miners toward fewer, larger pools — concentrating control over transaction selection.

Without Fast Relay

Block propagation takes seconds via the regular P2P network. TCP handshakes, serialization overhead, and geographic distance add up. Miners far from the block source waste hashrate on stale work. Orphan rate increases with distance.

With FIBRE

Block data reaches the other side of the planet in tens of milliseconds — approaching the speed-of-light bound in fiber optic cable. The latency advantage of large, well-connected pools is dramatically reduced. Geographic location matters far less.

The orphan rate equation is straightforward: if a block takes 5 seconds to propagate and blocks arrive every 600 seconds on average, roughly 0.83% of all mining work is wasted on stale blocks. Reduce propagation to 50ms and that drops to 0.008%. That difference directly impacts which miners remain profitable and which are squeezed out.

Core Innovation

How FIBRE Works

UDP + Forward Error Correction + Compact Blocks

▾

FIBRE is a patchset applied to Bitcoin Core that replaces the standard TCP-based block relay with a purpose-built UDP protocol optimized for minimum latency. Three key technologies work together to make sub-100ms global propagation possible:

UDP Transport

TCP requires a three-way handshake before data flows, and its congestion control algorithms throttle throughput. FIBRE uses UDP to start sending block data immediately — no waiting for acknowledgments. Lost packets are handled at a higher level by FEC, not by TCP retransmission.

Forward Error Correction (FEC)

Instead of waiting for lost packets to be detected and retransmitted (adding round-trips), FIBRE sends redundant coded data alongside the block. The receiver can reconstruct the full block even if some packets are lost in transit. This eliminates the need for retransmission round-trips entirely.

Compact Blocks (BIP 152)

Most transactions in a new block are already in the receiver's mempool. Compact blocks send only short transaction IDs — the receiver reconstructs the full block locally. This reduces the data that needs to travel over the wire from ~1-2 MB to just a few KB.

Step 1

New Block Found

Miner discovers valid block

→

Step 2

Compact + FEC Encode

Compact block + redundant data

→

Step 3

UDP Blast

Sent immediately, no handshake

→

Step 4

Reconstruct

FEC + mempool = full block

The combination is powerful: compact blocks minimize the data sent, FEC eliminates retransmission delays, and UDP removes connection setup overhead. The result is propagation that approaches the theoretical speed-of-light limit through fiber optic cable. FIBRE is distributed as a patchset that anyone can compile and run against Bitcoin Core, connecting to public or private relay nodes.

The FIBRE Network

Global nodes, real-world propagation statistics

▾

The resurrected FIBRE network consists of geographically distributed nodes running the patched Bitcoin Core client. Nodes connect to both public relay infrastructure and the regular Bitcoin P2P network, ensuring blocks propagate through the fastest available path. Anyone can join by compiling the FIBRE patchset and connecting to the network.

~30ms

Avg Propagation

Global Nodes

<100ms

Worst Case Target

~3KB

Typical Block Data Sent

Node distribution spans key mining regions: nodes are positioned to minimize the maximum latency between any two points on the network. The public dashboard at stats.bitcoinfibre.org provides live propagation metrics.

US East NA

US West NA

Europe EU

Asia-Pacific APAC

South America SA

Additional Nodes GLOBAL

2013

Bitcoin Relay Network by Matt Corallo — the original fast relay, using TCP and optimized serialization. Reduced propagation from 10+ seconds to roughly 1-2 seconds.

2016

FIBRE v1 launched — Corallo introduces UDP + FEC. Propagation drops to tens of milliseconds. Ran successfully for several years.

2019-2024

FIBRE goes dormant. The original network winds down. Compact blocks (BIP 152) deployed in Core improves baseline P2P relay, but gap remains vs. dedicated relay.

2025

FIBRE resurrected. New maintainers update the patchset for modern Bitcoin Core, deploy fresh infrastructure, and launch the public stats dashboard. The network is live and accepting connections.

Why This Matters for Bitcoin

Mining decentralization is censorship resistance

▾

Bitcoin's censorship resistance ultimately depends on mining being sufficiently decentralized. If a small number of pools control the majority of hashrate, they can be coerced into censoring transactions. Today, the top 3-4 pools control well over 50% of hashrate. Network latency is one of the structural forces pushing miners toward these large, well-connected pools.

Levels the Playing Field

A solo miner in rural Australia and a pool operator in Virginia both receive new blocks at nearly the same time. Geography stops being a competitive disadvantage. Smaller pools and solo miners can compete on more equal footing.

Reduces Orphan Cost

Faster propagation means fewer orphans across the network. This isn't just good for decentralization — it reduces wasted energy and improves the economic efficiency of mining overall.

Complements Other Efforts

FIBRE works alongside Stratum V2 (miner-chosen transactions), compact blocks in Core, and Erlay (efficient tx relay). Each reduces a different centralization vector. Fast block relay specifically addresses the propagation latency vector.

The argument is economic: if block propagation is slow, miners pay a "latency tax" proportional to their distance from other miners. Rational miners minimize this tax by colocating with the largest pools or joining them directly. FIBRE makes this tax negligible for everyone, removing one of the structural incentives toward centralization. It doesn't solve mining centralization alone, but it eliminates one of the quieter, more insidious forces driving it.

FIBRE is Infrastructure, Not a Product

FIBRE is a public good for Bitcoin's decentralization. It's a patchset anyone can compile and a network anyone can join. The resurrection means this critical infrastructure — dormant for years while mining centralization continued to worsen — is operational again. The source code is available, the nodes are running, and the stats dashboard is public. If you run mining infrastructure, connecting to the FIBRE network is one of the simplest things you can do to contribute to decentralization.