From Materials Simulation to Experimental Astronomy, New NVIDIA AI Software Unlocks Scientific Discoveries

2026-06-22 · Source: NVIDIA Blog · Field: Science & Research — Artificial Intelligence & Machine Learning, Space Science & Astronomy, Physical Sciences & Chemistry · Depth: Advanced, short

Summary

NVIDIA is introducing new AI software at the ISC conference to accelerate scientific discovery in chemistry, materials science, and astronomy. The NVIDIA DAQIRI library, NVIDIA ALCHEMI NIM microservices, and the upcoming NVIDIA cuPhoton reference code transform CPU-intensive tasks into real-time, GPU-accelerated pipelines. CuPhoton, for example, accelerated FITS image loading and reading from the Rubin Observatory's LSST by 14,900x and signal processing by 8,400x using 32 NVIDIA Grace Blackwell superchips. DAQIRI streams data from fast detectors, enabling real-time AI analysis of collision data at CERN's ATLAS Experiment, preventing the loss of over 99% of signals. ALCHEMI microservices, including batched geometry relaxation (BGR) and batched molecular dynamics (BMD), accelerate chemical and materials discovery, simulating millions of molecules. Lila Sciences achieved a 50x speedup in materials screening and a 30% acceleration for magnetic property calculations using ALCHEMI, with TensorNet kernels providing a 6x speedup in training/inference and 3x memory reduction.

Key takeaway

For research scientists and engineers working with large-scale scientific data or complex simulations, NVIDIA's new software stack offers significant performance gains. You should evaluate cuPhoton for accelerating astronomical data analysis or DAQIRI for real-time sensor data processing. If you are in materials science, consider ALCHEMI microservices to dramatically speed up molecular simulations and high-throughput screening, potentially reducing weeks of work to days and enabling novel discoveries.

Key insights

NVIDIA's new GPU-accelerated software dramatically speeds scientific data processing and simulation across diverse research domains.

Principles

• GPU acceleration transforms scientific workflows.
• Real-time data processing prevents signal loss.
• Batched simulations enable high-throughput discovery.

Method

The software suite integrates high-performance networking (DAQIRI), petabyte-scale data processing (cuPhoton), and domain-specific microservices (ALCHEMI) to create end-to-end accelerated pipelines for scientific research.

In practice

• Accelerate FITS data analysis for astronomy.
• Perform real-time AI on detector collision data.
• Simulate millions of molecules for materials discovery.

Topics

• NVIDIA CUDA-X
• Scientific Computing
• Materials Discovery
• Experimental Astronomy
• Real-time AI
• GPU Acceleration

Code references

• NVIDIA/daqiri
• NVIDIA/nvalchemi-toolkit
• nvidia/megatron-lm
• NVIDIA/nvalchemi-toolkit-ops

Best for: AI Scientist, Research Scientist

Related on AIssential

• Accelerating AI-Powered Chemistry and Materials Science Simulations with NVIDIA ALCHEMI Toolkit-Ops — NVIDIA ALCHEMI Toolkit-Ops accelerates atomistic simulations by providing GPU-optimized, batched operations for chemistry and materials science.
• Making Sense of the Early Universe — AI and GPU acceleration dramatically enhance astronomical discovery by automating complex data analysis.
• Using Accelerated Computing to Live-Steer Scientific Experiments at Massive Research Facilities — GPU-accelerated computing enables real-time analysis and steering of exascale scientific experiments, drastically reducing data processing times.
• Accelerating science with AI and simulations — AI is at a second inflection point, integrating language and multi-modal reasoning to accelerate scientific discovery.
• How NVIDIA's Earth-2 uses AI to Accurately Predict Weather — NVIDIA Earth-2 provides an open, GPU-accelerated AI stack for faster, more accurate weather and climate forecasting.
• AI galaxy hunters are adding to the global GPU crunch — New space telescopes are generating unprecedented data volumes, necessitating GPU-accelerated deep learning for astrophysical analysis.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by NVIDIA Blog.