High-Performance Computing (HPC) in the cloud propels medical research ahead, facilitating advancements in healthcare.
In the past, high-performance computing (HPC) was restricted to researchers and institutions with access to massive HPC clusters. For those without direct access, the process involved writing proposals and waiting their turn, with limited resources and many people ahead in line. This often meant long wait times and potential repeats if mistakes were made or codes didn't run well in the large cluster environment.
All that has changed in recent years, thanks to the cloud. The cloud has revolutionized data processing, making HPC clusters more accessible to a wider range of researchers without the lengthy wait times. The AWS Summit Washington, DC 2025 featured Jianjun Xu, AWS Higher Education Research's principal solution architect, who discussed how Amazon Web Services (AWS) has streamlined the research lifecycle, delivering quick results that propel healthcare forward. Joseph Marcotrigiano, the chief of the structural virology section at the National Institutes of Health (NIH), shared how the agency uses AWS tools to better understand cardiovascular disease and uncover new treatment options.
If you're intrigued, click here to read the latest CDW Cloud Computing Research Report.
AWS on Demand HPC Resources for Healthcare Researchers
On-premises clusters do have their benefits, but establishing a traditional HPC cluster can take up to seven or eight months! By the time an organization acquires the necessary hardware and sets it up, the technology might already be outdated. Add to that the difficulty of procuring graphics processing units (GPUs), and using HPC services through AWS seems like a no-brainer. AWS ensures that organizations have immediate access to the latest hardware, according to Xu.
AWS offers various HPC options for healthcare organizations. The AWS Parallel Computing Service is a fully managed SLURM cluster, allowing researchers to create a personalized cluster based on their processor types and latency needs within 20 minutes. Users can control the compute nodes and build their own node groups. Moreover, users can run native apps or containerized apps on AWS with the SLURM scheduler.
"You can create a compute environment running up to 100,000 CPUs, but you'll only be charged for what you use," said Xu. "It's on demand."
An alternative service is AWS ParallelCluster, geared towards researchers who want full control of the SLURM scheduler and its plug-ins. It's an open-source solution that lets users create a customized HPC cluster in the cloud that they manage themselves.
Researchers can choose from over 800 HPC instance types, alongside resources like Amazon FSx for Lustre and Amazon File Cache to support HPC goals.
"We don't want you to waste any resources, so you only pay for what you use," said Xu.
NIH and HPC: Uncovering New Treatments for Cardiovascular Disease
Cardiovascular disease is the number one cause of human mortality worldwide. In 2019, approximately 18.6 million people died from the disease globally. High levels of low-density lipoprotein in the blood raise the risk of cardiovascular disease, as these particles can accumulate in the blood and form plaques on artery walls, potentially leading to heart attacks or strokes.
In the United States, 30% to 40% of the population aged 50 and older takes statins to manage high cholesterol. Statins target the receptor, not the particle itself. To learn more about the particle itself, scientists at the NIH recently employed HPC and cryo-electron microscopy to model the LDL particles. Marcotrigiano mentioned that, until recently, this level of detail was considered impossible.
The model creation required massive amounts of data—35,000 movies and about 17.5 terabytes to be precise. The movies had to be compressed into high-resolution images, and researchers then exhaustively aligned particles based on similarities and differences, classifying particles using both 2D and 3D systems.
As a result, researchers now have a better understanding of how the particle binds to the receptors, which will aid in the development of new therapies that target the particle directly rather than just the receptors.
"The only place we could do this was in the cloud," said Marcotrigiano, adding that the NIH utilized Amazon FSx for Lustre and several GPUs to process and store the data for this project.
Follow these steps to successfully deploy high-performance computing.
Additional Topics
- Cloud Computing
- Data Analytics
- Data Management
- Hardware
Associated Articles
- The onset of advanced technology and data-and-cloud-computing has transformed cardiovascular-health research by making high-performance computing (HPC) accessible and efficient, accelerating discoveries in medical-conditions like cardiovascular disease.
- With AWS Parallel Computing Service, healthcare researchers can quickly create and control customized HPC clusters using the latest hardware, ensuring efficient processing of large datasets, such as those required for cryo-electron microscopy, and uncovering new data-analytics opportunities in science.
%20in%20the%20cloud%20propels%20medical%20research%20ahead%2C%20facilitating%20advancements%20in%20healthcare.){kind=link}