Milky way evolution revealed through AI-enhanced star-by-star simulation

Galactic disk of gas (Image: RIKEN)

Scientists have simulated the Milky Way using 100 billion stars. The project models galaxy evolution across 10,000 years accurately. Researchers achieved this using artificial intelligence and supercomputers combined. This breakthrough offers unprecedented insight into galactic dynamics and formation.

Who Made This Possible?

The research team was led by Keiya Hirashima in Japan. RIKEN, University of Tokyo and Barcelona’s Universitat collaborated closely. They combined astrophysics expertise with high-performance computing and AI.

Seven million CPU cores worked simultaneously to perform the calculations. AI accelerated simulations, reducing run-time dramatically compared to traditional methods.

When Was This Simulation Completed?

The breakthrough was achieved using modern supercomputers in 2025. One million years of galactic evolution now takes only 2.78 hours. This reduces simulations that previously took decades to mere months.

Where Could This Technology Be Applied?

Beyond astrophysics, AI-assisted simulations may aid climate and ocean studies. The weather forecasting, environmental monitoring and multi-scale scientific research benefit similarly. This approach opens possibilities across multiple global research domains efficiently.

What the Simulation Shows? 

The simulation tracks each star’s orbit, birth and eventual death. It also models supernova explosions and gas expansion in detail. Researchers can now study chemical element distribution across galaxies.

Galaxy dynamics can be examined at an unprecedented resolution level. This helps scientists test theories about our Milky Way’s structure.

What This Means for Science?

The study proves AI can go beyond simple pattern recognition. It enables discovery in large-scale, multi-physics simulations across disciplines. Future research can explore previously unreachable scales of space and time.