NVIDIA unveiled three new software tools at ISC Hamburg — cuPhoton for astronomical data processing, DAQIRI for streamin

NVIDIA introduced new software at the ISC conference in Hamburg this week designed to accelerate artificial intelligence for scientific research across chemistry, materials discovery and dark matter exploration. The tools — NVIDIA DAQIRI library, NVIDIA ALCHEMI NIM microservices, and NVIDIA cuPhoton reference code — convert work that traditionally required hours or days on CPUs into real-time GPU-accelerated workflows. These products are part of NVIDIA CUDA-X, a broader collection of tools and libraries for AI and high-performance computing.

cuPhoton is a reference code that enables scientists to extract insights from multidimensional data collected by telescopes, X-ray instruments and laser experiments. It accelerates loading, processing, analysis and visualization of petabytes of data. Princeton University collaborated with NVIDIA to develop cuPhoton, and both Princeton and Harvard plan to use it for processing data from observatories and dark energy surveys. In early testing, cuPhoton achieved a 14,900x speedup in loading and reading FITS images from the Rubin Observatory's Legacy Survey of Space and Time, using NVIDIA GB200 NVL72 systems. Signal processing and analysis using 32 NVIDIA Grace Blackwell superchips was up to 8,400x faster.

NVIDIA DAQIRI — Data Acquisition for Integrated Real-time Instruments — is a high-performance networking library for streaming data from fast detectors and sensors. Conventional systems drop data when instruments produce it faster than the system can save it. DAQIRI handles incoming streams in real time. The A-GHOST project, developed by scientists from CERN, the University of Chicago and University College London, uses DAQIRI to run AI analysis on collision data recorded by the ATLAS Experiment at CERN, analyzing data that would normally be discarded, capturing signals that would otherwise be lost.

NVIDIA ALCHEMI is a collection of domain-specific microservices and toolkit for accelerating chemical and materials discovery, with applications in battery materials, catalysts, OLED displays and beauty products. NVIDIA released two ALCHEMI NIM microservices in March for batched geometry relaxation and batched molecular dynamics, allowing researchers to simultaneously simulate millions of molecules and materials. An upcoming ALCHEMI NIM microservice for the Vienna Ab initio Simulation Package enables materials simulations with higher GPU throughput, delivering a 3x speedup for geometry optimization. An upcoming ALCHEMI Toolkit lets developers build custom atomistic simulation workflows and accelerate machine learning interatomic potentials. Lila Sciences, building an autonomous lab for life sciences and materials science, collaborated with NVIDIA using ALCHEMI to accelerate high-throughput materials screening by 50x, accelerate magnetic property calculations by 30%, achieve a 6x speedup in training and inference for TensorNet, and reduce memory usage by 3x, enabling simulations that previously took weeks to complete in days.