---
format: "aidr-story-markdown/v1"
id: "e21b1a157dc5cfe4be4b69dbe6fc36a6cf7ee60eb2c7cedee60ad89298721719"
canonical_url: "https://aidr.today/e21b1a15?lang=en"
title: "DeepSeek Elastic Compute (DSec): Sandbox Infrastructure for Effective Agentic Training at Scale"
lang: "en"
requested_lang: "en"
available_langs: ["en","vi"]
translation_fallback: null
fallback_fields: []
published_at: "2026-09-23T00:51:04.000Z"
category: "Models"
topics: ["anthropic","claude","multi-agent","open-source"]
source_urls: ["https://arxiv.org/abs/2609.22978","https://lobste.rs/s/3hbty3/deepseek_elastic_compute_dsec_sandbox"]
summary: "Large-scale agentic training and evaluation with large language models (LLMs) rely on isolated, stateful execution environments in which models inspect repositories, invoke tools, execute commands, and interact with task-specific services. These workloads create sandboxes in large bursts, span heterogeneous functionality and isolation requirements, retain state across long interactions, and draw from large image corpora with limited reuse. Supporting them therefore requires an elastic execution platform rather than a single sandbox runtime. This report presents DeepSeek Elastic Compute (DSec), a production sandbox platform that exposes FnCall, container, microVM, and full-VM sandbox backends through a unified SDK. DSec coordinates placement and lifecycle management across the cluster, composes environments from independently versioned layers, combines memory sharing, reclamation, and CPU scheduling for high-density execution, and loads image data on demand from Fire-Flyer File System (3FS), a cluster-wide distributed filesystem. DSec is co-designed with the reinforcement learning (RL) framework, decouples stateful rollout execution from preemptible GPU training, coordinates sandbo…"
---

# DeepSeek Elastic Compute \(DSec\): Sandbox Infrastructure for Effective Agentic Training at Scale

> [Open the canonical story](<https://aidr.today/e21b1a15?lang=en>)

**Published:** 2026-09-23T00:51:04.000Z
**Category:** Models
**Topics:** anthropic, claude, multi\-agent, open\-source

## Summary

Large\-scale agentic training and evaluation with large language models \(LLMs\) rely on isolated, stateful execution environments in which models inspect repositories, invoke tools, execute commands, and interact with task\-specific services\. These workloads create sandboxes in large bursts, span heterogeneous functionality and isolation requirements, retain state across long interactions, and draw from large image corpora with limited reuse\. Supporting them therefore requires an elastic execution platform rather than a single sandbox runtime\. This report presents DeepSeek Elastic Compute \(DSec\), a production sandbox platform that exposes FnCall, container, microVM, and full\-VM sandbox backends through a unified SDK\. DSec coordinates placement and lifecycle management across the cluster, composes environments from independently versioned layers, combines memory sharing, reclamation, and CPU scheduling for high\-density execution, and loads image data on demand from Fire\-Flyer File System \(3FS\), a cluster\-wide distributed filesystem\. DSec is co\-designed with the reinforcement learning \(RL\) framework, decouples stateful rollout execution from preemptible GPU training, coordinates sandbo…

## Sources

- [Story source](<https://arxiv.org/abs/2609.22978>)
- [Discussion](<https://lobste.rs/s/3hbty3/deepseek_elastic_compute_dsec_sandbox>)