SHREDS, EXPLAINED

From block production to recoverable data.

01

Fragment

BLOCK DATA

D1

D2

D3

D4

C1

C2

C3

C4

A leader fragments block data into data shreds and generates parity shreds for forward error correction.

02

Propagate

LEADER → ROOT

RAIDEN

PEERS

SIMPLIFIED NETWORK VIEW

Turbine distributes shreds through a tree of validators. Raiden receives the stream and delivers it to regional endpoints.

03

Recover

RECEIVE DATA + PARITY

D1

D2

?

C1

RECOVERABLE DATA

ENOUGH SHREDS REQUIRED

FEC parity can recover missing data when enough shreds arrive. Reconstruction happens in your stack with Surge, and at Raiden with Pulse.

Illustrative packet counts and topology. FEC is not an unlimited loss guarantee. Raw reception, reconstruction, transaction deserialization and on-chain confirmation are separate stages.

TWO PRODUCTS / DIFFERENT INTEGRATION WORK

Raw control or a reconstructed stream?

WHAT YOU RECEIVE

SURGE

PULSE

Data format

Raw shred bytes

Binary transaction payloads

Transport

UDP + TCP

gRPC streaming

Reassembly + FEC

Your pipeline

Handled by Raiden

Payload deserialization

Client-side

Client-side

Account filtering

Your pipeline

Server-side filters

Monthly / region

$499

$700

Free trial / region

3 hours daily

3 hours daily

Data stage

Pre-confirmation

Pre-confirmation

Published P99 / FRA

0.78 ms

1.32 ms

Choose Surge

Custom infrastructure, validator-grade integrations and low-level shred processing.

Explore Surge ↗

Choose Pulse

Launch detection, arbitrage applications, real-time indexers, analytics and flow analysis.

Explore Pulse ↗

P99 figures are vendor-published on raiden.wtf/shreds for a 100K-transaction Frankfurt comparison. The page does not identify peers or run date. Pulse removes client-side deshredding, not binary deserialization.

Start where the signal starts.

Start your free trial