Shanghai Starts the Mingzhu Plan for AI Compute in Orbit
Shanghai has launched the Mingzhu Plan, a long-term effort to develop a space-based computing constellation that organizers say could eventually reach gigawatt scale.
RSS feed for Chips & Compute →
Shanghai has launched the Mingzhu Plan, a long-term effort to develop a space-based computing constellation that organizers say could eventually reach gigawatt scale.
Two Beijing research institutions have used their second-anniversary event to present eight artificial-intelligence projects spanning industrial software, genomics, materials, molecular design, low-precision inference, physical intelligence, and single-cell biology.
Beijing has upgraded a Shijingshan facility that collects data for humanoid robots, relaunching it under the name Shijingshan Embodied Intelligence Training Center. The first phase of the site, previously known as the Shijingshan Humanoid Robot Data Training Center, was presented in its updated form during the China International Fair for Trade in Services.
China’s AI infrastructure debate has been dominated by models, chips, and data centers. A new report from the China Telecom Research Institute argues that the next pressure point will be token consumption as AI agents execute more tasks and generate far more inference traffic.
China Mobile has introduced a product called AI Trusted Computing, or AITC, as Chinese companies look for ways to run AI workloads on sensitive data without treating security as an afterthought. The launch took place on September 12 at the 2026 China Computing Power Conference in Langfang.
Huawei has demonstrated a new optical-interconnect design it says can move 7.2 terabits of data per second, putting a less visible part of AI infrastructure into the spotlight. Huawei unveiled the 7.2T NPO prototype at Shenzhen’s China International Optoelectronic Exposition on September 10.
Cambricon Technologies (中科寒武纪科技股份有限公司) says it completed Day-0 adaptation of DeepSeek-V4.1-Flash on the vLLM inference stack, allowing the model to run on Cambricon accelerators on the same day DeepSeek released it.
Enflame Technology’s first day as a public company gave China’s domestic AI-chip drive a visible market test. Shares in the Tencent-backed accelerator designer closed at 397 yuan on September 11, up 179% from their 142.18-yuan issue price, after opening at 410 yuan.
China’s AI hardware buildout is expanding beyond processors and memory into the components that move information and present it to users. Raysolve Semiconductor has completed a Pre-A4 financing round of nearly 100 million yuan, according to a September 10 report by 36Kr.
A Nanjing startup is betting that the next AI infrastructure challenge will be coordinating models rather than merely serving them. SinapisAI, also called Nanjie Intelligent, has raised an angel round of nearly 100 million yuan, according to a September 10 report by the Sci-Tech Innovation Board Daily carried by NetEase.
China’s push to run larger AI workloads outside centralized data centers is attracting a new class of chip funding. Shanghai-based Transcputing has completed a Series A round after raising nearly 2 billion yuan across six months, according to a September 10 report from Eastmoney and Shanghai Securities News.
China’s attempt to replace Nvidia hardware is running into a costly constraint that lies beside, rather than inside, the accelerator: high-bandwidth memory. Reuters reports that people familiar with nonpublic pricing discussions said Huawei and Cambricon had lifted indicated prices for current and planned AI processors as HBM became more expensive and harder to obtain.
Nano-Core Chip, also known as NCC, has announced a 1 billion yuan C1 financing round to support its work on 3D computing-in-memory chips. The Hangzhou company says it is developing a 3D-CIM architecture that combines DRAM-based near-memory computing, SRAM computing-in-memory, and RISC-V heterogeneous computing.
OpenBMB has released MiniCPM5-2B, a compact open-weights language model designed for local deployment, tool use, and agentic workflows. The model has 2,516,756,480 parameters, a 131,072-token context window, and an Apache-2.0 license, according to the OpenBMB model card. The release reflects a different direction from the race for ever-larger frontier models.
JD Cloud and Moore Threads have announced a plan that would test how far China’s domestic AI hardware can move beyond isolated deployments. At the 2026 JD Global Technology Explorer Conference on September 9, the partners said they intend to build a 100,000-card intelligent-computing cluster based on Moore Threads’ all-function GPUs.
DeepCtrls, a Nanjing company focused on AI control for energy and infrastructure, has reported a Series B+ investment in the several-hundred-million-yuan range, led by CATL. TMTPost reported the funding on September 8 and said Fosun Capital, Taiping Innovation Investment, and GF Xinde participated, with existing shareholders including Source Code Capital increasing their stakes.
Beijing Deep Space Matrix Technology has announced an angel financing round of more than 100 million yuan to support a planned space-AI computing business.
China’s integrated-circuit exports reached US$40.731 billion in August, up 129.8% from a year earlier, according to customs data reported by Nikkei Asia. The figure gives semiconductor trade an unusually prominent place in a broader export surge and offers another indication that global demand for computing hardware is shaping China’s industrial economy.
China has put a concrete number behind its next phase of AI infrastructure growth. A new Ministry of Industry and Information Technology plan sets 2030 capacity at 9,800 EFLOPS, a level that would require a sharp expansion from the 2,185 EFLOPS recorded at the end of June.
Hefei Zhongke Brain Intelligent Technology has announced a B+ strategic financing round worth hundreds of millions of yuan, with CRRC Capital leading and Ginkgo Valley Capital, Shuimu Fund, and TusPark Fund participating. The announcement was published through an authorized QbitAI release on September 7.