[Doc]: Best Practice for HICache

[hzh0425]

Motivation

Modifications

Accuracy Tests

Benchmarking and Profiling

Checklist

• Format your code according to the Format code with pre-commit.
• Add unit tests according to the Run and add unit tests.
• Update documentation according to Write documentations.
• Provide accuracy and speed benchmark results according to Test the accuracy and Benchmark the speed.

[gemini-code-assist]

Summary of Changes

Hello @hzh0425, I'm Gemini Code Assist¹! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

This pull request introduces comprehensive documentation for SGLang's HiCache feature. The new guide outlines the architecture, benefits, configuration, and deployment best practices for HiCache, a hierarchical KV caching system designed to significantly enhance performance and efficiency for large language models in long-context and multi-turn conversational AI applications. This addition aims to provide users with clear instructions and insights into leveraging HiCache effectively.

Highlights

• Introduction to HiCache: This document introduces SGLang's HiCache, a three-tier hierarchical KV caching system that extends RadixAttention. It's designed to significantly improve performance for long-context and multi-turn conversation scenarios by intelligently managing KV caches across GPU memory, host memory, and external storage backends, addressing capacity bottlenecks.
• Real-World Performance Gains: The documentation highlights substantial performance improvements, including a 56% reduction in Time to First Token (TTFT) and a 2x inference throughput improvement for coding agents, and an 84% reduction in TTFT for general QA scenarios, demonstrating HiCache's effectiveness.
• Architecture and Components: An overview of HiCache's three-tier memory hierarchy (GPU, host, and storage) is provided, along with its key components: HiRadixTree for page referencing, a Cache Controller for data movement, pluggable Storage Backends, and an Optimized Data Plane with GPU-assisted I/O kernels.
• Supported Storage Backends: The document lists various production-ready storage backends such as DeepSeek 3FS (HF3FS), Mooncake, AIBrix KVCache, and NIXL, as well as HiCacheFile for development and testing purposes, showcasing the system's flexibility.
• Configuration and Deployment Guidelines: Detailed guidelines are included for configuring core HiCache parameters, optimizing memory layout, and setting prefetch policies. It also provides practical deployment examples for HF3FS and Mooncake, and demonstrates integration with PD Disaggregation.
• Custom Storage Backend Integration: Instructions are given for integrating new storage backends, requiring the implementation of get(key), exists(key), and set(key, value) methods, and registration with the HiCache BackendFactory, emphasizing the extensibility of the system.

Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point by creating a comment using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands on the current page.

Feature	Command	Description
Code Review	/gemini review	Performs a code review for the current pull request in its current state.
Pull Request Summary	/gemini summary	Provides a summary of the current pull request in its current state.
Comment	@gemini-code-assist	Responds in comments when explicitly tagged, both in pull request comments and review comments.
Help	/gemini help	Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here.

You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.

Footnotes

1. Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩

[gemini-code-assist]

Code Review

This pull request introduces a comprehensive guide on SGLang HiCache best practices. The documentation is well-structured and covers the motivation, architecture, configuration, and deployment of HiCache. My review focuses on improving the clarity and accuracy of the examples by pointing out potentially incorrect or internal model names that might confuse users. I've also suggested a minor formatting fix.

[ykwd]

I think we also need to update ‎docs/index.rst

[merrymercy]

also need to link it in the index

sglang/docs/index.rst

Lines 35 to 40 in 336e9a6

	:caption: Advanced Features
	advanced_features/server_arguments.md
	advanced_features/hyperparameter_tuning.md
	advanced_features/speculative_decoding.ipynb
	advanced_features/structured_outputs.ipynb

[hzh0425]

also need to link it in the index

sglang/docs/index.rst

Lines 35 to 40 in 336e9a6

	:caption: Advanced Features
	advanced_features/server_arguments.md
	advanced_features/hyperparameter_tuning.md
	advanced_features/speculative_decoding.ipynb
	advanced_features/structured_outputs.ipynb

added, thx. @merrymercy