Breakthrough alert: AWS’s new HPC service could fast-track the next big discovery

Amazon Web Services, Inc. (AWS), a subsidiary of Amazon.com, Inc. (NASDAQ: AMZN), has officially launched the AWS Parallel Computing Service, a revolutionary managed service aimed at enhancing scientific and engineering research through advanced high-performance computing (HPC) capabilities. The service simplifies the setup and management of HPC clusters, enabling users to run complex simulations and models at unprecedented scales on AWS’s robust cloud platform.

The AWS Parallel Computing Service streamlines the creation of HPC infrastructure by utilizing Amazon Elastic Compute Cloud (Amazon EC2) instances, specialized storage, and low-latency networking tailored for HPC workloads. It allows system administrators to configure clusters using the widely-recognized Slurm workload manager, known for its efficiency in handling extensive computational tasks. This innovation is expected to drive breakthroughs in drug discovery, genomic research, engineering design, and meteorological applications by enabling rapid scaling of simulations and model validations.

In addressing common HPC management challenges, AWS Parallel Computing Service offers several key benefits. It simplifies deployment with tools like the AWS Management Console, Command Line Interface (CLI), and Software Development Kit (SDK), facilitating a smooth setup process. The service significantly reduces operational overhead by automating updates and patching, thus minimizing manual intervention. Users can now scale simulations efficiently, addressing complex problems and driving innovation across various fields. Furthermore, the service supports a diverse range of users, from startups to large enterprises and national laboratories, enabling seamless migration of existing Slurm-based workflows to the cloud without extensive re-architecture.

The AWS Parallel Computing Service also benefits from AWS’s high-performance infrastructure, including specialized compute instances such as AMD, Intel, and Graviton-based options. These instances offer up to 65% better price-performance compared to traditional x86-based options, providing a cost-effective and efficient computing solution.

Several leading organizations and research institutions have already adopted the AWS Parallel Computing Service. Marvel Fusion, a Germany-based fusion energy startup, is leveraging the service to advance its sustainable energy technology. Maxar Technologies, a leader in space technology and satellite imagery, is using the service to enhance its data processing capabilities. RONIN, a digital infrastructure expert, is exploring the service’s potential to optimize its computational resources. The National Renewable Energy Laboratory (NREL) is utilizing the service to drive advancements in energy technology and sustainability.

AWS Parallel Computing Service is available in several regions, including US East (Ohio), US East (N. Virginia), US West (Oregon), Europe (Frankfurt), Europe (Stockholm), Europe (Ireland), Asia Pacific (Sydney), Asia Pacific (Singapore), and Asia Pacific (Tokyo). Future developments will include CloudFormation support, allowing customers to build and deploy HPC clusters using infrastructure as code, further enhancing flexibility and ease of management.

Ian Colle, Director of Advanced Compute and Simulation at AWS, highlighted the service’s transformative potential: “Developing a cure for a catastrophic disease, designing novel materials, advancing renewable energy, and revolutionizing transportation are problems that we just can’t afford to have waiting in a queue. Managing HPC workloads, particularly the most complex and challenging extreme-scale workloads, is extraordinarily difficult. Our aim is that every scientist and engineer using AWS Parallel Computing Service, regardless of organization size, is the most productive person in their field because they have the same top-tier HPC capabilities as large enterprises to solve the world’s toughest challenges, any time they need to, and at any scale.”