Resources

AWS Well-Architected Lenses

• High Performance Computing Lens: design principles and best practices for HPC workloads on AWS.
• Machine Learning Lens: for HPC teams running ML training on shared GPU clusters.
• Generative AI Lens: for large-model training and inference on accelerated compute.

Build with infrastructure as code

• AWS HPC Recipes: ready-to-use IaC recipes for HPC on AWS, including AWS PCS and ParallelCluster.
• AWSome Distributed AI: reference architectures and examples for running distributed AI and ML workloads on AWS PCS and ParallelCluster.

• Apollo Tyres accelerates engineering workflows with HPC on AWS (IND368)
• Mission-Ready HPC: From NOAA today to AI tomorrow (WPS205)
• Engineering at full throttle: How AWS powers Formula 1 (SPF108)
• Boeing's PLM cloud transformation (IND321)

• Running a complex design environment on AWS (MFG205)
• EC2 Nitro networking under the hood (NET402)
• Natera's genomic innovation: scale clinical diagnostic operations (HLS206)
• How Merck improves drug design with biological foundation models (HLS215)
• AWS-accelerated computing enables customer success with generative AI (CMP207)

• Accelerating innovation with high performance computing on AWS (MFG105)
• Cutting-edge AI with AWS and NVIDIA (CMP208)
• AWS Graviton: The best price performance for your AWS workloads (CMP334)

• Remote desktop and application streaming with NICE DCV (2019)
• HPC on AWS: Innovating without infrastructure constraints (2021)
• The future of HPC is looking a lot like ML (2021)
• HPC on AWS: Solve complex problems with pay-as-you-go infrastructure (2022)

When organizations make their data open on AWS, anyone can access and analyze it without moving or storing it themselves. The Registry of Open Data on AWS catalogs 100+ PB of cloud-optimized public datasets, including:

• gnomAD: the Genome Aggregation Database, with variant data from hundreds of thousands of exomes and genomes for population genetics and clinical research.
• NOAA GOES: near-real-time and archival imagery from the GOES-R series of geostationary weather satellites.
• Sentinel-2: global, high-resolution optical satellite imagery for earth observation and geospatial analysis.
• Common Crawl: a petabyte-scale corpus of web crawl data, widely used to train large language and machine learning models.

See a complete list of Public Datasets available in the AWS Cloud.

For research and HPC workloads that span clouds, AWS Interconnect (multicloud) provides private, high-speed, dedicated network connections between your Amazon VPCs and other cloud providers.

Learn more »

The AWS Cloud Credit for Research program supports researchers building publicly available, cloud-hosted science-as-a-service tools, running proof-of-concept tests, or training their communities to use the cloud for research workloads.

Learn more »

AWS makes cloud budgeting more predictable by waiving data egress fees for qualified researchers and academic institutions: the fees for data transfer out from AWS to the internet.

Learn more »