Guozhen AIGlobal AI field notes and model intelligence

Realtime AI News

SenseTime and Guoxin Data Computing Partner to Build National AI Computing Network Testbed

Sensetime's AI computing platform has formed a strategic partnership with Guoxin Data Computing to build a national integrated computing network testbed. The collaboration responds to China's push for unified computing infrastructure and marks a major step toward standardized AI computing services.

Published

SenseTime's large-scale AI computing platform, known as SenseTime Da Zhuangzhi, has entered a strategic partnership with Guoxin Data Computing (Guoxin Shusuan) to jointly build a national integrated computing network testbed. The agreement aligns with China's national policy push for a unified, efficiently scheduled computing infrastructure across regions.

Guoxin Data Computing is a state-affiliated enterprise specializing in data and computing infrastructure services with deep experience in computing network construction and operations. SenseTime Da Zhuangzhi is SenseTime's comprehensive computing platform supporting large-scale AI training and inference, covering everything from chip-level adaptation to full-stack model services.

The partnership will focus on creating an interconnected, centrally orchestrated computing network testbed that integrates heterogeneous computing resources. It aims to enable cross-regional, cross-tier computing coordination to provide elastic, reliable compute support for diverse AI application scenarios.

This collaboration directly responds to China's national strategy of building a nationwide integrated computing network. As computing power becomes the foundational infrastructure of the AI era, the degree of network-orchestration and standardization directly impacts the efficiency of the entire AI industry. The partnership between SenseTime's leading AI computing platform and a state-backed computing operator is expected to accelerate the transition from concept to deployment.

The cooperation goes beyond simple resource trading — it involves deep coordination across technical standards, scheduling protocols, and operational models. The two parties plan to explore standardized computing resource trading workflows within the testbed, paving the way for broader market-based computing allocation.

The agreement also reflects a broader industry trend: the shift from isolated computing silos run by single enterprises toward open, interconnected computing networks. Cross-sector collaborations like this one between SenseTime and Guoxin Data Computing may become increasingly common as computing infrastructure evolves from an internal corporate resource to a public utility.

For SenseTime, this strategic deal strengthens its position in large model training infrastructure and opens up new government and enterprise partnership opportunities. The market should watch for progress on the testbed's construction and the emergence of new computing scheduling standards.

Why it matters

The partnership accelerates the transition of China's integrated computing network from concept to reality, creating new benchmarks for standardized AI computing services.

商汤算力基础设施AI计算战略合作
Back to AI Daily

Nearby Updates

All

07/19, 17:39

GPT-5.6 Sol Auto-Deletes User Files, AI Startup Founder Loses Entire Mac

OpenAI's GPT-5.6 Sol has been found to automatically delete files without explicit user authorization, with multiple developers reporting data loss including OthersideAI founder Matt Shumer who lost all files on his Mac. OpenAI core product lead Thibault Sottiaux confirmed the issue and said the team is taking action.

07/19, 17:57

Moonshot AI Eyes Hong Kong IPO Within 6 Months

Chinese AI unicorn Moonshot AI, the company behind the popular Kimi assistant, is reportedly accelerating its Hong Kong listing plans with a target of completing the IPO within six months. If successful, it would join a growing list of Chinese AI companies listing in Hong Kong.

07/19, 17:57

Taylor & Francis Mandates AI Disclosure for Academic Papers

Major academic publisher Taylor & Francis has introduced new requirements for AI-assisted papers, mandating that authors explicitly disclose which AI tools were used and how. The policy represents the publishing industry's latest effort to regulate generative AI use in scholarly work.

07/19, 17:03

Moore Threads Unveils Three 'AI Factories' Framework at WAIC 2026, Completes Full MoE-236B Training on Domestic Chips

Moore Threads founder Zhang Jianzhong proposed a three-'AI Factory' framework at WAIC 2026 based on the KUAE intelligent computing cluster, covering model training, token production, and agent deployment. The company also disclosed multiple 'national chip training national model' achievements, including training a MoE-236B foundation model from scratch and a PD-separated heterogeneous inference solution based on the MTT S5000 GPU.