mlsanityData to 7 Oct 2026

Thousand-GPU Large-Scale Training and Optimization Recipe for AI-Native Cloud Embodied Intelligence Infrastructure

arXiv 2603.11101 ↗First posted 11 Mar 2026cs.RO cs.AI cs.DCcited by 3

Not on the map. Why?

Abstract (arXiv)

Embodied intelligence is a key step towards Artificial General Intelligence (AGI), yet its development faces multiple challenges including data, frameworks, infrastructure, and evaluation systems. To address these issues, we have, for the first time in the industry, launched a cloud-based, thousand-GPU distributed training platform for embodied intelligence, built upon the widely adopted LeRobot framework, and have systematically overcome bottlenecks across the entire pipeline. At the data layer, we have restructured the data pipeline to optimize the flow of embodied training data. In terms of training, for the GR00T-N1.5 model, utilizing thousand-GPU clusters and data at the scale of hundreds of millions, the single-round training time has been reduced from 15 hours to just 22 minutes, achieving a 40-fold speedup. At the model layer, by combining variable-length FlashAttention and Data Packing, we have moved from sample redundancy to sequence integration, resulting in a 188% speed increase; π-0.5 attention optimization has accelerated training by 165%; and FP8 quantization has delivered a 140% speedup. On the infrastructure side, relying on high-performance storage, a 3.2T RDMA network, and a Ray-driven elastic AI data lake, we have achieved deep synergy among data, storage, communication, and computation. We have also built an end-to-end evaluation system, creating a closed loop from training to simulation to assessment. This framework has already been fully validated on thousand-GPU clusters, laying a crucial technical foundation for the development and application of next-generation autonomous intelligent robots, and is expected to accelerate the arrival of the era of human-machine integration.

Key facts. Cited by 3 papers in the map (data through 7 Oct 2026). Builds on 25 of its 31 references that are in the map. First posted on arXiv 11 Mar 2026.

Builds on

25 of 31 references.

References · 25

  1. 2023GPT-4 Technical Reportcited by 18,132
  2. 2023RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Controlcited by 2,993
  3. 2024OpenVLA: An Open-Source Vision-Language-Action Modelcited by 2,777
  4. 2024π0: A Vision-Language-Action Flow Model for General Robot Controlcited by 2,297
  5. 2015High-Dimensional Continuous Control Using Generalized Advantage Estimationcited by 1,916
  6. 2023Mastering Diverse Domains through World Modelscited by 1,193
  7. 2025GR00T N1: An Open Foundation Model for Generalist Humanoid Robotscited by 1,167
  8. 2017Mastering Chess and Shogi by Self-Play with a General Reinforcement Learning Algorithmcited by 723
  9. 2024Mobile ALOHA: Learning Bimanual Mobile Manipulation with Low-Cost Whole-Body Teleoperationcited by 377
  10. 2025SimpleVLA-RL: Scaling VLA Training via Reinforcement Learningcited by 141
  11. 2024AI Agents Under Threat: A Survey of Key Security Challenges and Future Pathwayscited by 122
  12. 2023DeepSpeed-Chat: Easy, Fast and Affordable RLHF Training of ChatGPT-like Models at All Scalescited by 107
13 more
  1. 2025VideoVLA: Video Generators Can Be Generalizable Robot Manipulatorscited by 61
  2. 2025MuJoCo Playgroundcited by 57
  3. 2025RLinf: Flexible and Efficient Large-scale Reinforcement Learning via Macro-to-Micro Flow Transformationcited by 51
  4. 2025ChatVLA-2: Vision-Language-Action Model with Open-World Embodied Reasoning from Pretrained Knowledgecited by 49
  5. 2025Dexbotic: Open-Source Vision-Language-Action Toolboxcited by 12
  6. 2024Prepacking: A Simple Method for Fast Prefilling and Increased Throughput in Large Language Modelscited by 10
  7. 2024Hummer: Towards Limited Competitive Preference Datasetcited by 10
  8. 2025Ring-lite: Scalable Reasoning via C3PO-Stabilized Reinforcement Learning for LLMscited by 10
  9. 2026RL-VLA3: A Flexible and Asynchronous Reinforcement Learning Framework for VLA Trainingcited by 7
  10. 2025SHARP: Synthesizing High-quality Aligned Reasoning Problems for Large Reasoning Models Reinforcement Learningcited by 5
  11. 2025Robot Learning: A Tutorialcited by 5
  12. 2025How Social is It? A Benchmark for LLMs' Capabilities in Multi-user Multi-turn Social Agent Taskscited by 3
  13. 2025Learning Truthful Mechanisms without Discretizationcited by 1

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 3.

3 papers.

  1. 2026Pre-VLA: Preemptive Runtime Verification for Reliable Vision-Language-Action and World-Model Rolloutscited by 4
  2. 2026Sword: Style-Robust World Models as Simulators via Dynamic Latent Bootstrapping for VLA Policy Post-Trainingcited by 3
  3. 2026D-VLA: A High-Concurrency Distributed Asynchronous Reinforcement Learning Framework for Vision-Language-Action Modelscited by 1