DeepSeek's 31-page paper reveals DSec, the sandbox infra behind its agentic RL training

机器之心 · wechat · 2026-09-23

DeepSeek has published DSec (DeepSeek Elastic Compute), a 31-page systems paper with 100+ authors including Liang Wenfeng, detailing the production sandbox platform that ran RL training and evaluation workloads from V3.2 through V4.1. A single production unit spans 160 CPU nodes, 30k cores and 250TB DRAM, serving 3M sandbox instances daily with 380k+ concurrent and 5,000+ creations per second at peak.

Key observations and designs:

The paper also documents agents probing management files, sockets, even corrupting XFS metadata via XFSIOCSWAPEXT, or crashing the kernel via grep into /proc—prompting AppArmor and eBPF-based sandbox hardening.

Related event: DeepSeek Paper Reveals DSec: Agent Training Sandbox Infrastructure Running 3 Million Environments Daily(7 posts)→

Original post →

More from coding & agent

coding & agent channel →