# PowerEdge XE8640 GPU cluster for molecular dynamics simulations and machine learning computational vibrational spectroscopy of ion channels

> **NIH NIH R35** · UNIVERSITY OF DELAWARE · 2024 · $170,417

## Abstract

Abstract
This application is in the response to NOT-GM-24-021 administrative supplements for equipment purchases. It
will support acquisition of Dell’s PowerEdge Rack Server XE8640 equipped with state-of-the art 4th generation
of Intel Xeon Platinum processor, 4 NVIDIA H100 GPUs interconnected with NVLink technology, 1 Tb of RDIMM
RAM, and 30 Tb of SSD. This architecture will dramatically expand our group’s computational capabilities for
simulating voltage-triggered conformational transitions in the voltage-gating ion (K+) channels as proposed in
our R35 MIRA award. Our simulations use two copies of a KvAP protein embedded in a double bilayer system
(to control the potential) and solvated in water resulting in ca. 400,000 atoms. Simulation timescales required to
achieve hypothetical conformational changes can be as long as tens of microseconds at physiological voltages
or a few microseconds at elevated voltages. Therefore, a GPU acceleration is crucial to our progress. It will also
be integral to modeling of two-dimensional infrared spectra which is a point of collaboration with experimentalists.
We will be using machine learning models to correlate spectroscopic observables and structures. GPU-
accelerated machine learning is nowadays routine. We plan to use the XE8640 server for such simulations as
well. NVIDIA H100 has already been shown to be a powerful architecture for molecular dynamics and machine
learning.

## Key facts

- **NIH application ID:** 11036819
- **Project number:** 3R35GM150963-02S1
- **Recipient organization:** UNIVERSITY OF DELAWARE
- **Principal Investigator:** Alexei Kananenka
- **Activity code:** R35 (R01, R21, SBIR, etc.)
- **Funding institute:** NIH
- **Fiscal year:** 2024
- **Award amount:** $170,417
- **Award type:** 3
- **Project period:** 2023-07-15 → 2028-04-30

## Primary source

NIH RePORTER: https://reporter.nih.gov/project-details/11036819

## Citation

> US National Institutes of Health, RePORTER application 11036819, PowerEdge XE8640 GPU cluster for molecular dynamics simulations and machine learning computational vibrational spectroscopy of ion channels (3R35GM150963-02S1). Retrieved via AI Analytics 2026-05-24 from https://api.ai-analytics.org/grant/nih/11036819. Licensed CC0.

---

*[NIH grants dataset](/datasets/nih-grants) · CC0 1.0*