Skip to content
Franklin AI Logo
FRANKLINAI
ToolsNewsVideosPrompts
  • AI Benchmarks
  • AI Model Comparisons
  • AI Cash
  • AI Courses
  • AI Dana
  • AI Games
  • AI Glossary
  • AI Model Types
  • Agent Skills
  • MCP Servers
  • AI GitHub Repos
  • AI Products
  • AI Research
  • AI Utilities
  • Chrome Extensions
  • Code Tests
  • Codex Pets
  • Collections
  • Image Tests
  • Newsletter
  • NotebookLM
  • Ridiculous Benchmarks
  • Vibes
Login
FRANKLINAI
AI Tools AI News AI Videos AI Prompts
About Franklin AI
Login / Sign Up
Search
  1. Home
  2. ›Learn
  3. ›AI Glossary
  4. ›Model sharding
CoursesResearchGlossaryModel TypesUtilities
Browse all 509 AI glossary terms
Topic
Hardware & infrastructure

Updated 7 November 2026

Share

AI Glossary

Model sharding

What does “Model sharding” mean?

Model sharding: Splitting a model's weights or computation across multiple devices.

Model sharding visual guide: Splitting a model's weights or computation across multiple devices.
Visual guide ·Open full size (new tab)

What is an example of “Model sharding”?

A model too large for one GPU is loaded across several GPUs.

Keep learning

See also

GPUTermComputeTermQuantizationTermTensor parallelismTermPipeline parallelismTerm

Same topic

Related terms

NPUTermModel servingTermNVLinkTermModel registryTermOn-device AITerm

Get a Free AI Prompt Guide

Sign up once and get the free guide immediately, then keep up with AI through one useful weekly email.

About · Contact · Privacy · Terms