IBM, RIKEN and Cleveland Clinic simulate a 12,635-atom protein
May 5, 2026
Cleveland Clinic, RIKEN and IBM reported on May 5, 2026 that they simulated a 12,635-atom protein using quantum computers and supercomputers. The work matters for chemistry and drug research.
IBM and RIKEN combine quantum computers with supercomputers
Cleveland Clinic, RIKEN and IBM reported on May 5, 2026 that they had modeled a protein with 12,635 atoms. According to the announcement, IBM quantum computers and two of the world's powerful supercomputers were combined. This is not a new AI app, but it is a research signal for data-intensive science, chemistry and later AI-assisted drug discovery.
The system targets large molecules
The 12,635-atom scale is the central number in the announcement. Proteins of this size are difficult for classical simulation because many interactions have to be considered. The reported approach combines quantum and high-performance computing to model complex molecules more effectively.
AI benefits indirectly from better simulations
AI models in biology and chemistry need high-quality data and reliable physical approximations. If quantum and supercomputing methods describe larger molecules more accurately, later models and analysis tools can benefit. The important limit: the announcement is primarily quantum and HPC research, not proof of a new drug.
HPCwire frames the step as hybrid supercomputing
HPCwire reported the same day on a breakthrough in quantum-assisted supercomputing. That supports the view that the practical value starts in research infrastructure. For companies, the key question is when such methods move from labs into repeatable workflows for materials research or pharma.
Why it matters
Europe and the United States are investing heavily in AI, quantum computing and supercomputing because new materials and medicines create enormous computing workloads. Progress in hybrid simulation can improve the data foundation on which AI systems train or test hypotheses. The short-term benefit is more relevant for research institutions than for ordinary IT departments.
A practical example
A pharmaceutical research team in Basel evaluates 200 protein variants for an early drug hypothesis. Today it would combine AI screening, classical simulation and lab validation. In a few years, a hybrid quantum-HPC approach could help model the hardest candidates more precisely before expensive lab experiments begin.
💡 In plain English
Researchers rebuilt a very large protein inside computers. Later, this can help people understand diseases and new medicines better.
Key Takeaways
- →The announcement was published on May 5, 2026.
- →The modeled protein contains 12,635 atoms according to the announcement.
- →Cleveland Clinic, RIKEN and IBM are involved.
- →The first value lies in quantum, supercomputing and research infrastructure.