GalaxyVS: China's AI drug platform cuts screening from years to seconds
Synopsis
China's GalaxyVS platform, powered by Tianhe supercomputers and Tsinghua's DrugCLIP method, screens drug compounds in tens of seconds — one million times faster than the existing world supercomputing record — potentially transforming how quickly treatments for cancer, rare diseases and outbreaks can be found.
Key Takeaways
GalaxyVS , unveiled by the National Supercomputing Centre in Tianjin on 27 May 2026 , reduces initial drug compound screening from months or years to tens of seconds.
The platform's molecular docking throughput is six orders of magnitude — roughly one million times — faster than the current world supercomputing record.
GalaxyVS was built in collaboration with Tsinghua University's Institute for AI Industry Research and runs on China's new-generation Tianhe supercomputers .
The system integrates DrugCLIP , a virtual screening method published in the journal Science in January by the Tsinghua team.
Target disease areas include tumours , neurodegenerative conditions , rare diseases , emerging infectious diseases , and potential public health crisis scenarios.
China has launched GalaxyVS, an artificial intelligence-powered drug discovery platform that compresses the initial chemical compound screening phase from months or years down to tens of seconds, according to the National Supercomputing Centre in Tianjin, which announced the breakthrough on Monday, 27 May 2026. The system was developed in collaboration with Tsinghua University's Institute for AI Industry Research and is powered by China's latest generation of Tianhe supercomputers.
What GalaxyVS does
GalaxyVS is designed to screen vast libraries of chemical compounds to identify promising lead molecules — a task that traditionally demands years of laboratory work. The platform's daily throughput in predicting molecular docking interactions is six orders of magnitude higher than the existing world record for supercomputing molecular docking, meaning it operates roughly one million times faster than the current benchmark. Developers said they expect the system to open a new pathway for identifying drug candidates targeting tumours, neurodegenerative conditions, rare diseases, and emerging infectious diseases.
The DrugCLIP connection
Central to GalaxyVS is DrugCLIP, an ultra-fast virtual screening method developed by the Tsinghua University team and published in the peer-reviewed journal Science in January. By integrating DrugCLIP with the raw computational muscle of the Tianhe supercomputer infrastructure, researchers were able to achieve throughput levels previously considered out of reach for molecular docking simulations. The platform reportedly processes interactions between two or more molecules at a scale no existing supercomputing system has matched.
Why it matters
Drug discovery is one of the most resource-intensive processes in modern medicine, with lead identification alone often consuming years and hundreds of millions of dollars before a single candidate enters clinical trials. A platform capable of collapsing that timeline to seconds could fundamentally alter the economics of pharmaceutical research, particularly for diseases where speed is critical — such as during public health emergencies. Developers noted that GalaxyVS could also accelerate drug research during outbreak scenarios, potentially shortening the gap between pathogen identification and therapeutic response.
The competitive backdrop
China's push into AI-driven drug discovery places it in direct competition with US and European firms that have invested heavily in computational biology platforms. Global players including Google DeepMind, whose AlphaFold reshaped protein-structure prediction, and a growing cohort of biotech startups have been racing to automate early-stage drug research. The GalaxyVS announcement signals that state-backed supercomputing infrastructure is now being deliberately channelled into life sciences, not just defence or climate modelling.
What's next
The National Supercomputing Centre in Tianjin has not yet disclosed a timeline for making GalaxyVS available to pharmaceutical partners or academic institutions, nor has it detailed any commercial licensing arrangements. The platform's real-world validation will depend on whether the lead molecules it identifies translate into viable drug candidates in downstream wet-lab and clinical-trial stages. Observers will be watching whether China's domestic pharmaceutical sector moves quickly to integrate the tool — and whether international collaborations follow.
Point of View
Controlling the computational layer of drug discovery could prove as strategically significant as controlling semiconductor fabrication. What mainstream coverage tends to underplay is the publication angle: anchoring DrugCLIP in Science before the platform launch is a credibility play aimed squarely at international pharmaceutical and academic audiences. The harder question — whether speed in virtual screening translates to higher clinical success rates — remains unanswered and will define whether GalaxyVS reshapes the industry or remains an impressive benchmark.
NationPress
12 Jul 2026
Frequently Asked Questions
What is GalaxyVS and what does it do?
GalaxyVS is an AI-powered drug discovery platform developed by the National Supercomputing Centre in Tianjin in collaboration with Tsinghua University . It screens large libraries of chemical compounds to identify potential drug candidates, compressing a process that previously took months or years down to tens of seconds.
How much faster is GalaxyVS than existing drug screening methods?
GalaxyVS is reportedly one million times faster than the existing world record for supercomputing molecular docking, a difference of six orders of magnitude in daily throughput. This speed is achieved by combining China's new-generation Tianhe supercomputers with the DrugCLIP virtual screening method.
What is DrugCLIP and how does it relate to GalaxyVS?
DrugCLIP is an ultra-fast virtual screening method developed by Tsinghua University's Institute for AI Industry Research and published in the journal Science in January . It forms the core AI engine inside GalaxyVS , enabling the platform to predict molecular interactions at unprecedented speed when paired with Tianhe supercomputing infrastructure.
Which diseases could GalaxyVS help treat?
Developers said GalaxyVS is expected to help identify lead drug molecules for tumours , neurodegenerative conditions , rare diseases , and emerging infectious diseases . The platform may also accelerate drug research during public health crises, where speed from pathogen identification to therapeutic candidate is critical.
How does GalaxyVS compare to global AI drug discovery efforts?
GalaxyVS enters a competitive field that includes platforms such as Google DeepMind's AlphaFold and numerous biotech startups focused on computational drug discovery. What distinguishes the Chinese approach is the direct integration of state-owned supercomputing infrastructure, giving GalaxyVS raw processing capacity that privately funded rivals may struggle to match at equivalent cost.
