---
title: "Cerberus Whitepaper & Academic Validation"
url: "https://radix.wiki/contents/tech/research/cerberus-whitepaper"
version: "2.0.1"
updated: 2026-08-02
last_verified: 2026-08-02
license: CC-BY-4.0
license_url: "https://creativecommons.org/licenses/by/4.0/"
---

# Cerberus Whitepaper & Academic Validation

|  |  |
| --- | --- |
| **Title** | [Cerberus](/contents/tech/core-protocols/cerberus-consensus-protocol): Minimalistic Multi-shard Byzantine-resilient Transaction Processing |
| **Published** | Journal of Systems Research (JSys), June 2023 |
| **arXiv** | [2008.04450](https://arxiv.org/pdf/2008.04450) |
| **Status** | Peer-reviewed; the sharded braiding protocol it specifies has not been implemented in production |

## Overview

The [Cerberus](/contents/tech/core-protocols/cerberus-consensus-protocol) whitepaper presents the formal specification of the Cerberus consensus protocol. Originally published as a preprint in 2020, it was [peer-reviewed and published](https://www.radixdlt.com/blog/cerberus-consensus-peer-reviewed) in the **Journal of Systems Research (JSys)** in June 2023.

### Key Contributions

- **Braided cross-shard consensus** – dynamic consensus groups formed per transaction across the relevant shards
- **3-phase commit protocol** – pre-prepare, prepare, commit, with formal safety proofs
- **Minimalistic design** – simple per-shard BFT instances composed via a braiding protocol
- **Deterministic state provisioning** – enabling global consensus between shard groups

### What Has and Has Not Been Demonstrated

The paper's contribution is a formal specification with safety proofs, and that contribution stands on peer review. What has _not_ happened is an implementation of the sharded braiding protocol in a production system.

The [Hyperscale public test](/contents/tech/research/hyperscale-500k-tps) of January 2026 – 500,000+ sustained TPS with peaks above 700,000 across 128 shards, on real cross-shard swap transactions rather than simple transfers, with linear scaling confirmed as shard count doubled from 64 to 128 – **did not validate the braiding protocol**. It measured the Radix Foundation's Hyperscale implementation, which the lead developer of the Xi'an production candidate states ["never really used Cerberus"](https://t.me/hyperscale_rs). The live [Babylon](/contents/tech/releases/radix-mainnet-babylon) mainnet runs Cerberus unsharded, a configuration in which braiding is inactive, and the Xi'an candidate [hyperscale-rs](/contents/tech/research/hyperscale-rs) does not implement Cerberus at all.

### Criticism

The paper's terminology has drawn criticism from the Xi'an candidate's lead developer, who argues it uses ["shards" for what are really keys](https://t.me/hyperscale_rs/10358) – encouraging the mistaken idea that ledger state is "pre-sharded" rather than mapped to shards by a topology that changes as shards split and merge. He has also noted that the paper mentions a "global shard" for tracking network-wide state without designing one, a role the Xi'an candidate fills with an explicit beacon chain.

## External Links

- [Full Paper (arXiv PDF)](https://arxiv.org/pdf/2008.04450)
- [Cerberus Consensus Peer Reviewed (Blog)](https://www.radixdlt.com/blog/cerberus-consensus-peer-reviewed)
- [Consensus Evolution at Radix](/contents/tech/research/consensus-evolution)
- [hyperscale-rs – the Xi'an production candidate](/contents/tech/research/hyperscale-rs)