HaiAI123

Curated Global AI Tools Directory

Industry update
Flat

Huawei Cloud trains embodied world model on Ascend, opens training stack

11/20/2026 — 11/20, 17:20·1 sources (all)·1 reports (all)

Key takeaways

  • Huawei Cloud says it trained an embodied world model on its Ascend compute, reaching SOTA on two leaderboards, and plans to open the full tr…

Background & analysis

According to a QbitAI report published on 2026-11-20 at 16:04, Huawei Cloud has trained an embodied world model on Ascend compute, and the model has reached SOTA on two benchmarks. The same report states that Huawei Cloud also plans to open the complete training foundation to the public. The piece frames the shift as one driven by the maturing of domestic compute: competition in embodied AI, it argues, is moving away from raw chip performance and into a multi-dimensional contest over training efficiency, data scale, deployment cost, and the speed of real-world deployment. The report does not specify which two benchmarks the model leads, nor does it give technical details about the model itself. Details on when the training foundation will open, and how external parties will gain access to it, are likewise absent from the available material, so the story currently stops at the announcement of the training result and the open-access plan.

AI-generated from 1 reports · updated 2 days ago

updream
24-hour heatHeat index 29 · peak 58 · 23h ago
24 hours agonow

Reports on this story headlines open the original

Nov 20
  1. QuantumBit reports that Huawei Cloud trained an embodied world model on Ascend compute, claiming SOTA results on two leaderboards, and plans to open up the full training stack. The report frames maturing domestic compute as shifting embodied AI competition from raw chip performance to training efficiency, data scale, deployment cost and real-world rollout speed.

    量子位AI score 56

Other stories people are talking about

How is heat calculated?About the method

Heat counts how many independent sources covered a story in the last 48 hours: one source counts once no matter how many posts it published, decaying with a 24-hour half-life. What ranks first is what many people are talking about.

This page aggregates public feeds. Headlines and summaries are machine-organized and remain the property of the original authors; verify important facts at the source. If you believe a headline or summary infringes your rights, email the contact address in the footer with the page URL and basis of your claim, and we will remove or replace it after review.

Surge
Discussion rising fast
New
First report within 6 hours
Rising
Still gathering discussion

Back to AI industry updates →