explainx.ainewsletter3.5k
TrendingNewsPathwaysSkills
Pricing
explainx.ai

Upskill in AI — 16 free pathways, live workshops & bootcamps, and 50+ courses from practitioners. Plus the skills, tools, and MCP servers to practice on.

follow us

corporate training

support@explainx.ai

get started

Find your pathTake Free Evaluation

learn

pathways — start freeworkshopsbootcampscoursescertificationsmock testsexplainx universitycorporate traininglearn skills & mcp

discover

skillsmcp serversexplainx mcptoolsagentsllmsdesignsdictionaryagi trackerranks

company

aboutvisionmissionteaminstructorscommunityhackathonscareers

content

daily AI newsstate of AI — live resultsblogreleasespromptsgeneratorsresource libraryfor LLMsexplainx.ai kids

solutions

all solutionsdeveloper upskillingmarketing upskillingproduct manager upskillingleadership upskilling

newsletter · weekly

Get AI news, tools, and insights in your inbox.

supportcontactprivacytermsdata rightshow we create contentsubmission guidelines

© 2026 AISOLO Technologies Pvt Ltd

  1. Home
  2. /
  3. Dictionary
  4. /
  5. VRAM
Infrastructure & Hardwareaka GPU memoryaka video memory

VRAM

VRAM is the memory on a GPU that stores model weights, activations, and KV cache during inference or training — the main constraint on what model size fits on which GPU.

Ask Melo about this← all terms

Video RAM sits physically on the GPU board and provides the high-bandwidth data path that compute cores need. When a model is 'too large for a single GPU,' VRAM is usually the binding constraint: a 70B-parameter model in fp16 needs roughly 140 GB, exceeding any single consumer GPU. Techniques like quantization, offloading, and tensor parallelism exist primarily to work around VRAM limits.

Related terms

Graphics Processing UnitHigh-Bandwidth MemoryQuantizationMemory BandwidthTraining ClusterModel Observability