Motivation behind KVBM — Dynamo
Title: Motivation behind KVBM — Dynamo
Published Time: Fri, 22 Aug 2025 17:35:16 GMT
Markdown Content: Skip to main content
Back to top- [x] - [x]
Ctrl+K
DynamoSearch Ctrl+K
Search Ctrl+K
Table of Contents
Architecture & Features
Using Dynamo
- Writing Python Workers in Dynamo
- Disaggregation and Performance Tuning
- Working with Dynamo Kubernetes Operator
Deployment Guides
- Dynamo Deploy Quickstart
- Dynamo Cloud Kubernetes Platform
- Manual Helm Deployment
- Minikube Setup Guide
- Model Caching with Fluid
Examples
- Hello World
- LLM Deployment Examples using VLLM
- LLM Deployment Examples using SGLang
- Multinode Examples using SGLang
- Planner Benchmark Example
- LLM Deployment Examples using TensorRT-LLM
Reference
-
Motivation...
Motivation behind KVBM#
Large language models (LLMs) and other AI workloads increasingly rely on KV caches that extend beyond GPU and local CPU memory into remote storage tiers. However, efficiently managing the lifecycle of KV blocks in remote storage presents challenges:
-
Tailored for GenAI use-cases
-
Lack of visibility into real-time block usage patterns.
-
Need for lightweight, ownership-driven memory management over complex object stores with unneeded overheads.
-
Modular and need simplified UX and to be memory safe.
-
Inability to differentiate between hot (frequently accessed) and cold (infrequently accessed) blocks across the stack without intrusive application-level changes.
-
Difficulty in optimizing storage placement across heterogeneous storage tiers (for example, SSDs, object storage, and cloud storage).
Conventional systems either lack dynamic feedback mechanisms or require deep integration into core storage paths, which both increases complexity and reduces portability.
previous KV Block Managernext KVBM Architecture
Privacy Policy | Manage My Privacy | Do Not Sell or Share My Data | Terms of Service | Accessibility | Corporate Policies | Product Security | Contact
Copyright © 2025-2025, NVIDIA Corporation.
Links/Buttons:
- Skip to main content
- document.write(
<img src="../_static/nvidia-logo-horiz-rgb-wht-for-screen.svg" class="logo__image only-dark" alt="Dynamo - Home"/>); Dynamo - GitHub
- Support Matrix
- High Level Architecture
- Distributed Runtime
- Disaggregated Serving
- KV Block Manager
- Motivation
- KVBM Architecture
- Understanding KVBM components
- KVBM Further Reading
- KV Cache Routing
- Planner
- Pre-Deployment Profiling
- Load-based Planner
- SLA-based Planner
- Dynamo Architecture Flow
- Writing Python Workers in Dynamo
- Disaggregation and Performance Tuning
- Working with Dynamo Kubernetes Operator
- Dynamo Deploy Quickstart
- Dynamo Cloud Kubernetes Platform
- Manual Helm Deployment
- Minikube Setup Guide
- Model Caching with Fluid
- Hello World
- LLM Deployment Examples using VLLM
- LLM Deployment Examples using SGLang
- Multinode Examples using SGLang
- Planner Benchmark Example
- LLM Deployment Examples using TensorRT-LLM
- Glossary
- NIXL Connect API
- #
- Privacy Policy
- Manage My Privacy
- Do Not Sell or Share My Data
- Terms of Service
- Accessibility
- Corporate Policies
- Product Security
- Contact