CATEGORY · PAGE 5 OF 27

Chips & Compute

Thursday, September 24, 2026

MODELSCHIPS & COMPUTE
DeepSeek Builds a Sandbox Layer for Agent Training
DeepSeek has described a less visible part of the agent race: the infrastructure needed to let an AI system try tasks safely, repeatedly, and at very large volume.

CHIPS & COMPUTEBIG TECH
Alibaba Links Its Next AI Chip to a Much Bigger Qwen Roadmap
Alibaba’s September 22 announcements were not a single product launch. They sketched a tightly connected plan: a new domestic accelerator, a longer model-development path, and a large expansion target for the cloud infrastructure needed to support both.

GEOPOLITICSCHIPS & COMPUTE
A U.S. Cloud Curb Could Open a New Front in China’s AI Compute Fight
For years, U.S. chip controls have focused on where advanced processors are sold and who can receive them. The next pressure point may be different: whether a Chinese company can use that hardware remotely after it has been installed in a data center outside China.

FUNDINGCHIPS & COMPUTE
SiliconFlow Adds New Capital for China’s AI Inference Layer
China’s AI investment cycle is not only about foundation models and chips. It is also about the infrastructure that turns a model into a service: scheduling workloads across different hardware, adapting models to chips, and providing the tokens that developers and businesses consume.
CHIPS & COMPUTEAPPLIED AI
Hygon Takes Its Chip Road Map From Data Centers to Physical AI
China’s AI hardware race has often been described in terms of data centers, large model training, and the effort to replace restricted foreign accelerators. Hygon Information Technology is now pointing at a different layer of the market: the chips that sit closer to machines, cameras, factory equipment, and robots operating in the physical world.
CHIPS & COMPUTE
CXMT Moves Its Fifth-Generation DRAM Platform Into Mass Production
China’s leading domestic DRAM producer has announced a new step in the country’s effort to build a more capable memory-chip supply chain. ChangXin Memory Technologies, known as CXMT, says its fifth-generation DRAM technology platform has entered mass production and is now underpinning two new mobile-memory products. The key word is “says.”

MARKETSCHIPS & COMPUTE
China’s AI Earnings Boom Leaves a Large Group of Listed Companies Behind
China’s listed-company earnings picture is improving in the aggregate, but the gains remain sharply uneven. A Nikkei Asia review of roughly 5,400 companies on mainland exchanges, excluding financial firms, found first-half net profit of about 1.9 trillion yuan, or US$283 billion, up 22% year on year.
CHIPS & COMPUTEAPPLIED AI
Huawei Updates Campus Networking for Enterprise Agent Workloads
Huawei has set out a campus-networking update aimed at workplaces where multiple AI agents, employee devices, business applications, and physical spaces must operate across the same network. The company calls the product the Stellar AI Campus Network Solution and says it combines superfast wireless, full-scope security, and network autonomy.

FUNDINGCHIPS & COMPUTE
RoboTechnik’s Proposed Hong Kong Listing Targets $800 Million
Shenzhen-listed RoboTechnik Intelligent Technology is planning a Hong Kong listing that could seek about US$800 million, according to The Economic Times. The report attributes the plan to two people with knowledge of the matter.
CHIPS & COMPUTEMARKETS
CXMT Eyes NAND R&D Line Amid AI-Driven Memory Shortage
Reuters reports that ChangXin Memory Technologies, or CXMT, is preparing to enter the NAND flash market, citing three people with knowledge of the private plans. Reuters describes CXMT as a DRAM specialist, meaning that the reported initiative would extend beyond the company’s publicly identified memory focus.

FUNDINGCHIPS & COMPUTE
D-Robotics Raises US$400 Million for Its Chinese Robotics Chip Platform
D-Robotics has completed a US$400 million Series C round, giving the Chinese company fresh capital to expand its Sunrise chip portfolio and the software layers it says are needed to develop and deploy AI robots.
MODELS BIG TECH
Z.ai Says GLM Helped Build Its Over-100,000-Accelerator AI Inference Stack
Z.ai says it used a GLM-5.3-powered infrastructure agent to help take GLM-5.3-Flash from initial hardware adaptation to a production inference service running on more than 100,000 Chinese-made AI accelerators. The September 17 technical account is not a new model release.
CHIPS & COMPUTEBIG TECH
Huawei Advances Ascend 960DT Timeline With an Up-to-4,096-NPU AI SuperPoD
Huawei has moved its Ascend 960DT schedule forward to the first quarter of 2027 while setting out its most detailed public specifications yet for an up-to-4,096-NPU computing system.

CHIPS & COMPUTEBIG TECH
Huawei Releases 3D Data Center Architecture at AIDC Wuhu 2026
Huawei announced a 3D Data Center architecture at the 2026 AIDC Industry Development Conference in Wuhu, Anhui, on September 15. Chinese media reported the release on September 16, with Guangming Daily noting the key elements highlighted by the company. Huawei characterizes the design as the world’s first 3D Data Center architecture.
CHIPS & COMPUTEFUNDING
Senwan Raises 100 Million Yuan for AI-Chip Power Delivery Components
Suzhou Senwan Electronics Technology Co., Ltd. (苏州森丸电子技术有限公司) completed a 100-million-yuan financing round labeled Series A in reports published September 15 and 16. The round was reported as co-led by Glory Ventures and an unnamed participant from the optical-module industry, with Shunrong Capital and Jiarui Capital participating, according to 36Kr.

CHIPS & COMPUTEMARKETS
China’s New Electronics Plan Puts AI Hardware at the Center of 2030 Goals
China has published a new five-year development plan for electronic-information manufacturing that ties the country’s industrial ambitions directly to advanced computing, integrated circuits, operating systems, and AI-related hardware.

CHIPS & COMPUTEBIG TECH
China Mobile Combines GPUs and Neuromorphic Chips for AI Inference
China Mobile Cloud and a group of chip, software, and academic partners have introduced a domestic GPU-plus-neuromorphic mixed-inference system, presenting it as a way to divide large-model inference across different processor types rather than relying only on conventional GPU clusters.
CHIPS & COMPUTEBIG TECH
Empyrean Uses AI Agents to Shorten China’s Chip Design Cycle
Empyrean Technology Co., Ltd. is putting a sharper metric on China’s push to bring AI agents into chip-design software.
CHIPS & COMPUTEMODELS
Chinese Researchers Advance a Tougher Ferroelectric Memory Material
A Chinese-led research team has reported a major endurance gain for a ferroelectric material being explored for future memory technologies, a development that could matter for high-performance computing and AI hardware over the longer term.

APPLIED AICHIPS & COMPUTE
Beijing Opens 100 Million Yuan AI Compute Program for Biomedicine
Beijing’s healthcare AI pilot base has announced a biomedicine-manufacturing support program that will provide more than 100 million yuan in computing resources for drug discovery, preclinical research, clinical trials, and associated data work. The initiative was unveiled at a September 12 CIFTIS healthcare forum and reported by CNR on September 13.