An increasing number of customers are choosing to run their high performance computing (HPC) workloads on AWS, because of the scale and flexibility the cloud offers. With access to virtually unlimited capacity, engineers, researchers and scientists can maximize their research efforts and reduce time to results when compared to a fixed on-premises HPC environment. HPC system owners can deploy a cloud-based cluster environment in minutes, incorporating a range of AWS services designed specifically for demanding HPC applications. This flexibility, and along with the on-demand nature of the cloud, allows for a different approach to meeting the needs of HPC users and applications.
To get the best performance for your workload, you should choose the right combination of services to meet the needs of your applications. AWS offers flexible compute, storage, networking and software to enable the creation of an HPC environment to match your specific requirements. A big part of this will be choosing the right Amazon Elastic Compute Cloud (EC2) instances from the wide range of EC2 instance types available.
Different families of EC2 instances are optimized for a range of workloads, allowing the selection of the correct CPU, GPU, FPGA, memory, storage and networking along with the correct operating system, based on the needs of your applications and the demands of your industry or vertical market.
To assist with identifying the key features and capabilities of these EC2 instances and then matching them to the appropriate workloads, this whitepaper identifies a number of popular HPC workloads and the most appropriate EC2 instances for each one. This will ensure you have the best match of compute platform to support your HPC requirements from the broad selection available in EC2.
To learn more about how to select the right EC2 instance for your HPC workload, read this whitepaper.