China’s Pujiang Forum Unveils Top 10 Sci-Tech Breakthroughs

China’s flagship science and technology gathering, the Pujiang Innovation Forum, has released its annual list of ten “hardcore” (硬核) achievements oriented toward the future — a showcase spanning artificial intelligence, next-generation computing, advanced semiconductors, fusion energy, and brain-computer interfaces. The list was announced at a dedicated achievements release event on September 11 in Shanghai, ahead of the main forum that opened September 12, as reported by People’s Daily.

The selections are more than a technology parade. They reflect a deliberate strategic pivot: China is pushing to move from laboratory “zero-to-one” breakthroughs toward “one-to-one-hundred” commercial and industrial scaling, with Shanghai positioning itself as a hub for the “future industries” (未来产业) that officials believe will define the next decade.

Four Domains, Ten Achievements

The list is organized around four clusters. Under AI applications, the Science and Technology Daily’s “Xiaoke Hotline” (小科热线) is a science-focused Q&A agent built on 40 years of the newspaper’s authoritative archives, designed to avoid the hallucination and authority problems of commercial AI products. Alongside it sits Hp-CARE, a gastric-cancer prevention and control ecosystem from Shanghai Xinchao Biotechnology that deploys China’s first Helicobacter pylori-specific large model to connect screening, diagnosis, treatment, and follow-up for grassroots clinics.

In the new computing category, the entries reach further. Zhangjiang Laboratory’s “Zhangjiang Guangqing” is an optoelectronic hybrid computing digital twin platform that lets researchers optimize algorithms and predict performance without tying up physical hardware — with the explicit ambition of building a software ecosystem as influential as NVIDIA’s CUDA, but for optical computing. The “Mingzhu Constellation Plan,” led by Shanghai Dongfang Tiansuan, aims to build a gigawatt-class space-based computing infrastructure, extending intelligent computing into low Earth orbit through a phased path from technical verification to scaled networking.

The quantum entry comes from Shanghai Buchou Quantum Technology: a large-scale atomic array quantum computing system, the “Liangchou-1” atomic quantum AI base model, built on programmable neutral-atom arrays with more than 1,500 qubits. Fusing a neutral-atom QPU with GPUs and classical CPUs, the system is described as the first internationally to achieve large-scale operation of neutral-atom quantum reservoir computing, targeting financial forecasting, dynamic risk assessment, and extreme weather warning.

Semiconductors, Fusion, and the Brain

The advanced packaging and processes cluster includes work from Yuanji Micro Technology on a standard process library and prototype devices based on 2D atomic-layer semiconductors. The company launched a 4-inch MoS2 wafer 2D prototype chip and opened foundry tape-out services, filling a domestic gap in 2D integrated-circuit process standardization.

Perhaps the most striking medical entry is Surugiolucel (舒瑞基奥仑赛注射液) from CARsgen Therapeutics — described as the world’s first approved CAR-T product for solid tumor treatment. It offers the first effective cell therapy option for patients with advanced gastric or gastroesophageal junction adenocarcinoma who have failed at least second-line therapy, with long-term survival beyond 50 months reported in first-line sequential trials. It was approved for market in June 2026. As CARsgen’s VP of business development Zang Peng noted, the milestone is not an endpoint: the company is already pursuing universal and in-vivo CAR-T approaches to overcome the cost and cycle-time limits of today’s autologous model.

In fusion energy, Shanghai Xinghuan Juneng Technology’s CTRFR-1 “Xinghuan-1” repetitive reconnection spherical tokamak device pursues a domestically original, globally unique technical route, fusing spherical tokamak high-beta advantages with magnetic reconnection heating. The device is planned for completion around 2028, with a target of equivalent Q>1 fusion performance in 2029–2030 and a foundation for a demonstration reactor in the early 2030s. Founder and CEO Chen Rui framed the company’s supply-chain relationship as two-way empowerment, using the device to continually raise new demands that pull supply-chain upgrades forward.

Two entries anchor the brain-computer interface category. Lingang Laboratory’s “Youren 1.0” ecosystem links seven research institutions and enterprises into a closed-loop “data flywheel,” including China’s first BCI-specific data-sharing platform and an implantable Chinese brain-sentence decoding framework spanning individuals and tasks. Shanghai Shishi Technology’s Rigi series is described as the world’s first brain-inspired sensing-computing integrated neural signal acquisition chip, using an event-driven architecture to deliver millisecond-level EEG readouts with 10–100x better energy efficiency than traditional solutions.

A Shared Research Paradigm Shift

If the achievements list is the hardware, the main forum supplied the philosophy. Under the theme “Sharing Innovation, Shaping the Future Together,” leading scientists converged on how AI is reshaping the scientific method itself. Shanghai AI Laboratory director Zhou Bowen argued that science increasingly needs AI to expand exploration, while AI needs science to provide real training grounds and standards — proposing that “scientific research is the next programming.” CAS academician Ding Hong observed that deep fusion of extreme-condition experimental devices with intelligent algorithms is sharply shortening the distance from experiment to discovery.

But Lingang Laboratory’s BCI chief scientist Li Chengyu offered the sharpest reframing, telling Yicai at the release event: “This is an era where computing power and capital are no longer scarce. What is truly scarce is not computing power, but the ability to ask questions.” A theme echoed across the forum was that young startups are increasingly building not just products but platforms — data foundations, process standards, supply chains, and ecosystems.

International cooperation was another refrain. Benoit Dubuis, a foreign academician of the Chinese Academy of Engineering and president of the Swiss Academy of Engineering Sciences, cited Switzerland’s openness and integration as the source of its success, and highlighted two unique advantages for cooperation with China: speed and scale. As Science and Technology Daily summarized the emerging consensus, the transformation of research paradigms in the AI era is no solo performance by any single discipline or nation, but a shared undertaking for the global scientific community.

What to Watch

Several of the claims — world’s-first CAR-T for solid tumors, the first brain-inspired sensing-computing chip, and the unique tokamak route — bear watching as independent verification and commercial deployment unfold. Two details warrant scrutiny: reporting has differed on whether Yuanji Micro’s 2D prototype uses a 4-inch or 8-inch wafer platform, and the timeline for the Xinghuan-1 fusion device stretches well into the 2030s.

What is clear is the direction of travel. Shanghai is betting that the next round of industrial advantage will be won not by single inventions but by the platforms, standards, and supply chains built around them — and it is assembling that ecosystem in public, one forum at a time.