Skip to content
Blockchain LabBlockchain LabTHE VENTURE STUDIO

LibraryConsensus2007Design paperCorpus record

Zyzzyva: Speculative Byzantine Fault Tolerance

Zyzzyva. Ramakrishna Kotla, Lorenzo Alvisi, Mike Dahlin, Allen Clement and Edmund Wong.

Replicas execute the leader's request speculatively and reply. If every replica agrees, the client is done. A disagreement starts a slower path that looks more like PBFT.

A reading of the public document. Not a copy of it, and not a claim about a later network that reused the name.

A finality gadget that shows the user a result before a quorum certificate should say so, the way this paper does.
Evidence
Primary paper
Re-measured
No
Assumptions
3
Records linked
2

Code

Source, repository, and what is actually hosted

  • SOSP 2007https://www.cs.utexas.edu/~lorenzo/papers/kotla07zyzzyva.pdf
  • The source URL is not a GitHub repository. No code remote is invented.
  • Full text is not mirrored. Public availability is not a licence to copy the file onto this site.
  • Add to a reading list

01 Claim ledger

What the paper is allowed to say

Each row is a sentence already in the study. The status is the same on every row: a model claim, not a live measurement.

  1. Claim 01 · Paper model

    The defect

    PBFT pays three rounds even when no one is faulty, which is almost always.

  2. Claim 02 · Paper model

    The rule

    Replicas execute the leader's request speculatively and reply. If every replica agrees, the client is done. A disagreement starts a slower path that looks more like PBFT.

  3. Claim 03 · Paper model

    How it is put together

    The fast path needs all replicas, not merely a quorum. The client is part of the protocol. It notices disagreement. Speculation means the reply can be ahead of a committed certificate.

  4. Claim 04 · Paper model

    Where the claim stops

    A fast path that needs every replica stalls when one is slow.

02 Three cuts

Observation, model, falsifier

A desk does not stop at the summary. Each claim is cut three ways, using only this study's own assumptions and checks. Nothing here is a new figure.

  1. 01 The defect

    Observation

    What the study says

    PBFT pays three rounds even when no one is faulty, which is almost always.

    Model

    What has to hold

    A fast path that needs every replica stalls when one is slow.

    Falsifier

    What would retire it

    Does the fast path require every replica?

  2. 02 The rule

    Observation

    What the study says

    Replicas execute the leader's request speculatively and reply. If every replica agrees, the client is done. A disagreement starts a slower path that looks more like PBFT.

    Model

    What has to hold

    A fast path that needs every replica stalls when one is slow.

    Falsifier

    What would retire it

    Does the fast path require every replica?

  3. 03 How it is put together

    Observation

    What the study says

    The fast path needs all replicas, not merely a quorum. The client is part of the protocol. It notices disagreement. Speculation means the reply can be ahead of a committed certificate.

    Model

    What has to hold

    A fast path that needs every replica stalls when one is slow.

    Falsifier

    What would retire it

    Can an executed request be undone?

  4. 04 Where the claim stops

    Observation

    What the study says

    A fast path that needs every replica stalls when one is slow.

    Model

    What has to hold

    A fast path that needs every replica stalls when one is slow.

    Falsifier

    What would retire it

    Does the fast path require every replica?

03 Sequence

One action, as an operating tape

  1. 01The leader orders a request.
  2. 02Replicas execute and reply to the client.
  3. 03The client accepts if the replies match. Otherwise it demands a view change or a commit proof.
  4. 04Does the fast path require every replica?

04 Load-bearing

The argument, and where a pitch drops it

  1. Why it is still on the desk

    The cut

    A finality gadget that shows the user a result before a quorum certificate should say so, the way this paper does.

    Why it carries weight

    If this cut is skipped, the paper's name is being used without the mechanism that makes the name mean anything.

    Where it is dropped

    The paper is crash-and-Byzantine in a known set, not a chain.

  2. After the text

    The cut

    Speculative BFT became a standard idea. Production chains usually wait for a certificate because rollback of a public ledger is the whole problem.

    Why it carries weight

    If this cut is skipped, the paper's name is being used without the mechanism that makes the name mean anything.

    Where it is dropped

    Speculative execution can be rolled back. The application has to allow that.

05 Register

What has to be true

  • Model · Not re-measured

    A fast path that needs every replica stalls when one is slow.

  • Stake and threshold · Not re-measured

    The paper is crash-and-Byzantine in a known set, not a chain.

  • Model · Not re-measured

    Speculative execution can be rolled back. The application has to allow that.

06 Divergence

What happened after the paper

Speculative BFT became a standard idea. Production chains usually wait for a certificate because rollback of a public ledger is the whole problem.

A later client, parameter set, or reward formula is a different object. Cite this paper for the mechanism. Cite a primary release for the network. This desk has not re-run the proof.

07 Pre-mortem

What to check before you use the idea

  4. 0 of 3 marked on this browser. A mark is a reading note, not a pass, a rating, or a recommendation.

08 Anatomy

The paper, in the order a builder needs

The problem it names

PBFT pays three rounds even when no one is faulty, which is almost always.

What the design proposes

  • The fast path needs all replicas, not merely a quorum.
  • The client is part of the protocol. It notices disagreement.
  • Speculation means the reply can be ahead of a committed certificate.

How the mechanism is specified

  • The leader orders a request.
  • Replicas execute and reply to the client.
  • The client accepts if the replies match. Otherwise it demands a view change or a commit proof.

What this page does not treat as proven

  • A fast path that needs every replica stalls when one is slow.
  • The paper is crash-and-Byzantine in a known set, not a chain.
  • Speculative execution can be rolled back. The application has to allow that.

Why the desk still reads it

A finality gadget that shows the user a result before a quorum certificate should say so, the way this paper does.

09 Lexicon

Terms, opened into the record

Speculative execution
Doing the work before a commit certificate exists.
Complete quorum
A reply set that includes every replica.

10 Repository

Every linked record on this page

Underlined words open a page that already exists: a concept, a protocol profile, a failure record, or another paper. If a word is not underlined, this desk does not have a record for it.

Concepts

This is Blockchain Lab's reading of a public design paper. It is not the paper, not a copy of it, and not an offer of tokens, equity, custody or a partnership. Later network behaviour can diverge from the text. Nothing here is investment, legal or technical advice.

Research status: Design paper. Last reviewed: 1 October 2026. This is a reading of a public paper, not investment, legal or security advice.