CHINA’S chip industry is breaking new ground on multiple fronts. The advances point to something bigger than a string of individual breakthroughs: a growing capacity to innovate across the semiconductor supply chain.
The industry momentum was evident at the 7th AI Computing Conference in Beijing on Monday, where next-generation chip architectures, computing-in-memory and 3D chip integration took center stage. The focus reflected a broader push across China’s chip industry to unlock greater computing power and efficiency as artificial intelligence (AI) demand surges, analysts said.The conference came amid advances by Chinese chipmakers across design, manufacturing, equipment and computing architectures, even as US semiconductor restrictions continue to tighten.On Sunday, Chinese memory-chipmaker CXMT announced mass production of its fifth-generation DRAM platform and unveiled two 24-gigabit LPDDR5X chips with higher capacities than their predecessors, another advance in a sector dominated by a handful of global players.Earlier, Huawei unveiled an accelerated road map for its Ascend AI processors, with the next-generation Ascend 960DT expected to be ready in the first quarter of 2027. The company also outlined plans for annual chip upgrades through 2029, alongside advances in processor interconnection and computing architecture.Xiang Ligang, a veteran telecommunications industry expert, told the Global Times on Monday that “China’s semiconductor development is not about a breakthrough in one particular field or technology. It is about comprehensive development across the entire industrial chain.”He pointed to progress in equipment, chip design, fabrication, packaging and materials, saying that sustained investment since 2021 had helped China establish capabilities in areas where domestic alternatives were previously limited.“China’s homegrown semiconductor technologies have begun catching up with world-class standards in most areas, and the momentum is unstoppable,” Xiang said.China has made significant progress in high-end chip design, but advanced manufacturing remains a key area where it lags behind global leaders, Liu Gang, chief economist at the Chinese Institute of New Generation AI Development Strategies, told the Global Times.“Improvements in domestic fabrication over the past two years have also helped companies such as Huawei advance their high-end processors and AI chips, providing stronger manufacturing support for China’s chip design capabilities,” Liu said.Liu attributed the industry’s progress to a combination of long-term technological accumulation, sustained research investment and policy support, including national research programs, industrial investment funds and local government initiatives.He cited Huawei’s “Tao’s Law,” which emphasizes innovation in chip architecture as a path to greater computing power, as an example of how years of research and technological accumulation are beginning to yield breakthroughs across multiple areas.Xiang highlighted China’s vast domestic market, infrastructure, talent pool and financial resources as key advantages. “Few countries can match China’s combination of strengths, which provides a solid foundation for advances across the semiconductor industry,” he said.China’s expanding semiconductor industry is also reflected in trade. Integrated circuit exports reached $216 billion in the first seven months of 2026, up 99.5 percent year on year and surpassing the full-year 2025 total of $201.9 billion, according to customs data.Market research firm Omdia recently sharply raised its 2026 growth forecast for China’s semiconductor market, projecting that it will reach $812.08 billion this year, with year-on-year growth of 92.9 percent.Domestic chipmakers are also deepening their integration with China’s AI industry. GPU developer Moore Threads, for example, has announced adaptations of its computing platform to models developed by DeepSeek, Alibaba’s Qwen and Zhipu AI. Such efforts are broadening the applications of homegrown computing hardware as China’s AI industry expands.Despite recent progress, sustaining the momentum will require further advances in both chip architecture and manufacturing technology. Liu identified computing-in-memory and specialized architectures tailored to AI applications as important directions, particularly as demand grows for more efficient computing in areas such as autonomous driving.“Having an innovative chip architecture alone is not enough if manufacturing capacity cannot keep up,” he said, stressing the need for innovation across the industrial chain.
