Parallel Universe
  • Learn
    • Introduction to PUT
    • Getting started with PUT
  • Architecture
    • What is a PUT Cluster?
    • Clusters
      • PUT Clusters
      • RPC Endpoints
      • Benchmark a Cluster
      • Performance Metrics
    • Consensus
      • Synchronization
      • Leader Rotation
      • Fork Generation
      • Managing Forks
      • Turbine Block Propagation
      • Commitment Status
      • Secure Vote Signing
      • Stake Delegation and Rewards
    • Validators
      • Overview
      • TPU
      • TVU
      • Blockstore
      • Gossip Service
      • The Runtime
  • CLI
    • Command-line Guide
    • Install the PUT Tool Suite
    • Command-line Wallets
      • Command Line Wallets
      • Paper Wallet
      • File System Wallet
      • Support / Troubleshooting
    • Using PUT CLI
    • Connecting to a Cluster
    • Send and Receive Tokens
    • Staking
    • Deploy a Program
    • Offline Transaction Signing
    • Durable Transaction Nonces
    • CLI Usage Reference
  • Developers
    • Get Started
      • Hello World
      • Local development
      • Rust program
    • Core Concepts
      • Accounts
      • Transactions
        • Overview
        • Versioned Transactions
        • Address Lookup Tables
      • Programs
      • Rent
      • Calling between programs
      • Runtime
    • Clients
      • JSON RPC API -1
      • JSON RPC API -2
      • JSON RPC API -3
      • Web3 JavaScript API
      • Web3 API Reference
      • Rust API
    • Writing Programs
      • Overview
      • Developing with Rust
      • Deploying
      • Debugging
      • Program Examples
      • FAQ
    • Native Programs
      • Overview
      • Sysvar Cluster Data
    • Local Development
      • PUT Test Validator
    • Backward Compatibility Policy
  • Validators
    • Running a Validator
    • Getting Started
      • Validator Requirements
    • Voting Setup
      • Starting a Validator
      • Vote Account Management
      • Staking
      • Monitoring a Validator
      • Publishing Validator Info
      • Failover Setup
      • Troubleshooting
    • Geyser
      • Geyser Plugins
  • Staking
    • Staking on PUT
    • Stake Account Structure
  • Integrations
    • Add PUT to Your Exchange
    • Retrying Transactions
  • Library
    • Introduction
    • Token Program
    • Associated Token Account Program
    • Memo Program
    • Name Service
    • Feature Proposal Program
    • NFT Program
      • Overview
      • Interface
      • Usage Guidelines
        • Create a new NFT-Mint
        • Cast NFT
        • Transfer an NFT
        • Change account status
        • Permission settings
        • Query Interface
        • Continuous casting
        • Change the Mint attribute
      • Operation Overview
        • Create a new NFT-Mint
        • Transfer NFT
        • Destroy
        • Freeze NFT accounts
        • Update
    • PUT multi-sign program
      • Overview
      • Interface
      • Usage Guidelines
        • Create a multi-signature account
        • Create a proposal account
        • Vote proposal
        • Verify Proposal
        • Add-singer
        • Remove-signer
      • Operation Overview
        • Create a multi-signature account
        • Create a proposal account
        • Vote
        • Verify
        • Add-singer
        • Remove-signer
  • PUT Privacy Policy
Powered by GitBook
On this page
  • Pipelining
  • Pipelining in the Validator
  1. Architecture
  2. Validators

Overview

PreviousValidatorsNextTPU

Last updated 2 years ago

Anatomy of a Validator

Pipelining

The validators make extensive use of an optimization common in CPU design, called pipelining.

Pipelining is the right tool for the job when there's a stream of input data that needs to be processed by a sequence of steps, and there's different hardware responsible for each.

The quintessential example is using a washer and dryer to wash/dry/fold several loads of laundry. Washing must occur before drying and drying before folding, but each of the three operations is performed by a separate unit.

To maximize efficiency, one creates a pipeline of stages.

We'll call the washer one stage, the dryer another, and the folding process a third. To run the pipeline, one adds a second load of laundry to the washer just after the first load is added to the dryer.

Likewise, the third load is added to the washer after the second is in the dryer and the first is being folded. In this way, one can make progress on three loads of laundry simultaneously.

Given infinite loads, the pipeline will consistently complete a load at the rate of the slowest stage in the pipeline.

Pipelining in the Validator

The validator contains two pipelined processes, one used in leader mode called the TPU and one used in validator mode called the TVU.

In both cases, the hardware being pipelined is the same, the network input, the GPU cards, the CPU cores, writes to disk, and the network output.

What it does with that hardware is different.

The TPU exists to create ledger entries whereas the TVU exists to validate them.