Realtime AI News
Envision's Ulanqab Xinghe base goes live with the world's largest AI computing 'super unit'
Envision Group announced on August 6 that its Ulanqab Xinghe base in Inner Mongolia has entered production, home to what it describes as the world's largest AI computing super unit. The roughly 120,000-square-meter facility targets million-card parallel computing at million-P scale, powered largely by direct green electricity, and is the flagship project of Envision's Gobi Mission.
Envision Group announced on August 6 that it has built what it calls the world's largest AI computing "super unit" in Ulanqab, Inner Mongolia, marking the official launch of its Envision Ulanqab Xinghe base. The facility, reported by Chinese tech outlet QbitAI, is a super computing hub built in the Gobi region, powered by a very high share of directly connected green electricity, and positioned as the first flagship project of Envision's "Gobi Mission."
In scale terms, the super unit spans about 120,000 square meters (roughly 20 standard soccer fields) and is designed for million-card parallel computing at million-P scale. Envision says it is the single AI data center with the strongest token output capacity in the world, setting a new record for AI infrastructure density.
The Xinghe base has a total planned capacity of more than 2GW and is set to become China's largest AI computing park and "token factory." As domestic AI computing moves toward 100,000-card and even million-card interconnects, the base is designed to host large-scale deployments of domestic AI chip clusters at a single site, offering a replicable blueprint for homegrown compute.
Power is the backbone of a million-card facility, and Envision's AI power system provides a dedicated "energy highway" and "power granary" for the super unit. The company says the base can deliver up to 10 times the compute output of a traditional data center of the same footprint, and will keep resetting global density records.
The Ulanqab location was chosen deliberately. The city is one of eight nodes in China's "East Data, West Computing" national project, giving it strong network latency advantages for real-time compute workloads, and the region's green electricity share is among the highest in the country at 67%. The base draws on locally built wind farms and dedicated transmission lines for direct green power supply.
Zheng Zihao, general manager of Envision's AIDC business, said Chinese chips are taking a path of cluster-based breakthroughs, and that the super unit provides domestic compute clusters with deployment space, power supply, and energy logistics — "the strongest assist for China's computing boom."
The "Gobi Mission" was formally unveiled at the VivaTech conference in France in June 2026, with a goal of building 5GW of green AI computing capacity in desert regions worldwide by 2030. The Ulanqab Xinghe base is the first step of that mission and a showcase for exporting Chinese AI data center solutions.
One caveat: the article was provided by Envision and republished by QbitAI with authorization. What to watch next is whether the super unit reaches full utilization, how effectively it serves domestic chip clusters, and how quickly follow-on Gobi Mission projects land outside China.
Why it matters
The launch of the world's largest AI computing super unit shows China scaling domestic chip clusters through a green-power direct-connection model, setting a template for AI data center construction in the million-card era.
Nearby Updates
All08/06, 13:36
MiniMax H3 tops open source community, defining a new bar for video models
MiniMax has open-sourced its new multimodal generation model H3, which ranks first on Artificial Analysis' video editing leaderboard and Arena's image-to-video chart, and has become the most popular model on Hugging Face. More than 100 domestic and international partners integrated H3 within 24 hours of its release, cementing it as a milestone for Chinese open-source AI expanding from language to video models.
08/06, 13:43
Meta discloses AI test breach, third such case after Anthropic and OpenAI
Meta has disclosed a breach involving its AI test content, becoming the third AI company to do so after Anthropic and OpenAI. Details about the scope and impact remain limited, but the pattern is drawing renewed attention to test-data security across frontier labs.
08/06, 14:00
OpenAI and the American Psychological Association launch three-year partnership on youth mental health and AI
OpenAI has announced a three-year partnership with the American Psychological Association (APA) to develop guidance, resources, and safeguards for responsible AI use supporting youth mental health. The collaboration marks one of the most direct efforts by a leading AI company to bring professional psychological expertise into AI safety and governance.
08/06, 12:40
Major critical safety flaws exposed in recent Anthropic and OpenAI AI safety tests, report says
A report from Chinese tech media 36Kr says major critical safety flaws were exposed in recent AI safety tests involving Anthropic and OpenAI. The claim raises fresh doubts about whether current safety evaluation methods can be trusted to gate frontier model deployment.