Just about all manufactured products are enabled by chemistry and materials science. Nevertheless, recent discoveries are costly and time-consuming and sometimes hindered by trial-and-error approaches. Computational methods akin to molecular dynamics with classical force fields traditionally struggle to predict chemical properties and stability accurately.
NVIDIA ALCHEMI (AI Lab for Chemistry and Materials Innovation) enables Chemistry and Materials Science researchers and developers with NVIDIA NIM microservices and toolkits optimized on NVIDIA accelerated computing platforms. This post introduces two recent atomistic simulation NIM microservices, the NVIDIA Batched Conformer Search (BCS) NIM and NVIDIA Batched Molecular Dynamics (BMD) NIM.
Accelerated conformer search and molecular dynamics using NVIDIA ALCHEMI
Molecules have multiple three-dimensional shapes, or conformers, each with distinct properties and stabilities. Identifying low-energy conformers is crucial for accurate chemical and material property prediction.
The ALCHEMI BCS NIM efficiently identifies and ranks low-energy conformers of molecules. Traditional quantum chemistry methods for conformer optimization are computationally expensive. BCS NIM uses AIMNet2 as a machine learning interatomic potential (MLIP) to speed up energy optimization, reducing the time required to generate low-energy conformers in comparison with quantum chemistry. This is useful for high-throughput conformer search campaigns, enabling researchers to rapidly evaluate thermally accessible molecule conformers at near quantum chemistry accuracy.
The BCS NIM works in three steps:
- The RDKit or nvMolKit library is used to generate multiple initial conformers from the SMILES format. The utmost variety of conformers could be specified or estimated based on the variety of rotatable bonds in a molecule.
- The conformers are then optimized using AIMNet2 asynchronously into their local energy minima. The optimizations proceed until the utmost force on any atom is below 0.005 eV/Å threshold.
- The ultimate step involves curating the conformer ensemble by validating bond connectivity based on interatomic distances, discarding high-energy conformers above a user-specified threshold, and deduplicating conformers based on relative differences in interatomic distances.
Figure 1 shows that as molecular complexity increases—with more rotatable bonds—conformer generation must also increase to make sure sufficient sampling of low-energy conformers. For instance, as much as 1,000 conformers could also be required for a molecule with 12 rotatable bonds to make sure all conformers are inside ≤3 kcal/mol of the lowest-energy structure. This shows the necessity for efficient conformer search methods. With the BCS NIM, average optimization time is 10-100 ms per conformer on an NVIDIA H100 GPU, for Fmax = 0.005 eV/Å.
In lots of workflows, conformer search is combined with molecular dynamics (MD) simulations. The BMD NIM for high-throughput simulations uses MLIPs akin to MACE-MPA-0, TensorNet, and AIMNet2 to handle this need. Use cases include high-throughput MD simulations of many atomistic systems and agentic AI workflows that require high-throughput asynchronous simulations.
Key features of the BMD NIM include:
- Dynamic batching: Optimize GPU utilization by dynamically batching atomic systems, for concurrent processing of multiple simulations to maximise throughput.
- GPU-based integrators: Perform simulations at a continuing variety of atoms, volume, and temperature (NVT), or a continuing variety of atoms, pressure, and temperature (NPT), using a Langevin thermostat and Monte Carlo barostat for temperature and pressure control.
- MLIP support: Support for MACE-MPA-0, TensorNet-MatPES-r2SCAN-v2025.1, TensorNet-MatPES-PBE-v2025.1, AIMNet2, AIMNet2-NSE, and AIMNet2-CPCM.
Figure 2 shows the performance gain (time per atom per step) as batch size increases, demonstrating the advantages of dynamic batching. Improved performance and bigger batch size are achieved with newer generations of NVIDIA GPUs, akin to the NVIDIA B200. For instance, 1.4 μs per atom per step and greater than 350,000 atoms simulated on one NVIDIA HGX B200 GPU with TensorNet. The benchmark is performed using batches of randomly sampled structures from the OMat24 dataset.
Accelerating OLED molecular and immersion cooling fluids discovery
NVIDIA customers are achieving groundbreaking results with ALCHEMI’s recent capabilities.
Universal Display Corporation (UDC) is accelerating OLED molecular discovery using ALCHEMI BCS and BMD NIM microservices. They’re predicting thermal processing stability as much as 10,000x faster than traditional density functional theory (DFT), while achieving near-comparable accuracy. UDC researchers can explore a much larger solution space, increasing the likelihood of finding best-performing OLED molecules.
Starting with a pool of candidate molecules, the researchers use the BCS NIM to discover and rank low-energy conformers. For every conformer, a simulation cell is constructed, and MD simulations are run to evaluate thermal processing stability. The flexibility to guage 1000’s of possibilities with unprecedented speed and accuracy was enabled by batching and MLIP. As an alternative of simulating each molecule individually, ALCHEMI executes quite a few simulations concurrently with the BCS NIM and BMD NIM, accelerating throughput and innovation.
One other collaborator, ENEOS, discovered over 1,000 promising environmentally friendly and non-PFAS immersion cooling fluids using a multi-stage workflow with the BCS NIM, BMD NIM, and the batched DFT microservice. An immersion cooling fluid needs a low dielectric constant to take care of signal integrity and a high flash point to scale back flammability. Starting with 10M molecules, they calculated dipole moments for 1M molecules using the BCS NIM. They used a knowledge-guided pre-training of graph transformer (KPGT) model to predict dipole moments for the remaining 9M, retaining only those with low dipole moments.
The batched DFT microservice computed and screened for molecules with low polarizability. Next, the BMD NIM was used to calculate bulk dielectric constants and discover candidates with low dielectric constants. Finally, flash point predictions were made using a KPGT model. This workflow enabled ENEOS to discover over 1,000 promising immersion cooling fluids with each low dielectric constant and high flash point in only three weeks.
Learn more
In the event you’re at Supercomputing 25, attend the ENEOS talk on the NVIDIA booth and take a look at the NVIDIA ALCHEMI demo.
Read NVIDIA Accelerated Computing Enables Scientific Breakthroughs for Materials Discovery and Accelerating the Way forward for Transportation with SES AI’s NVIDIA-Powered Innovation for Electric Vehicles and watch the ALCHEMI video.
Acknowledgments
Because of George Fitzgerald, Sean Ryno, Gbolade Kayode, and Darice Liu from Universal Display Corporation, Hideki Ono, Hiroyuki Tsujimoto, Kentaro Yomogita, Yoichiro Kawami, and Masanao Goto from ENEOS, in addition to Saee Paliwal and Srimukh Veccham from NVIDIA, for his or her contributions.
