Deploying Secure Container Solutions from Edge to Exascale

HPC Containers

Hardened Edge Containers

Cloud and Data Center Solutions

Singularity

Streamlining Workflows Innovating Security:

The Sylabs Advantage

Software containers have quickly grown in popularity in recent years for their ability to provide consistency across different environments, simplify deployment, enable scalability, isolate applications, and maximize resource efficiency.

Sylabs has a proven track record for using containers to solve big workflow and security challenges in performance-intensive and mission-critical environments. Singularity, our widely adopted container runtime, was specifically developed to support mobility of compute. It implements a unique security model through the Singularity Image Format (SIF) to mitigate privilege escalation risks and provides a platform to capture a complete application environment into a single Open Container Initiative (OCI) compliant file.

Solution Brief

Performance Portability for AI & ML:
The Role of Software Containers in the Specialized Hardware Era

This brief explores the challenges of the growing specialized hardware realm and how Singularity’s unique capabilities, backed by Sylabs’ expertise, stand out for professionals navigating the AI and ML landscape.

Use Cases

AI / Machine Learning

From training in the data center to inferencing at the edge, Sylabs makes it easy to manage your containerized workflows with machine learning frameworks and leverage the power of GPUs while optimizing security and repeatability.

Bioinformatics/ Computational Biology

Containerize genomics analyses and other workflows to provide long-term reproducibility and ease deployment to local HPC and the cloud.

Edge Security Hardening

Sylabs develops advanced tools that seamlessly integrate with existing DevSecOps workflows to shield mission-critical edge and IoT device software from IP theft.

Organizations today need the flexibility to securely package applications, libraries, and data sets in one scalable container, enabling interoperability and reproducibility. This allows for seamless movement across different systems or hardware environments without compromise, ensuring consistent behavior and performance.

$

At the edge

$

On-premise

$

In the cloud

Trusted by hundreds of organizations

ABCI
AIST
RIKEN

Singularity Containers 101

Elevate Your Computer Science Curriculum

Sylabs has developed the Singularity Containers 101 curriculum. This comprehensive resource will enrich the academic journey of computer science students by providing them with a foundational understanding of Singularity container technology and its role in high performance computing environments.

What Singularity Users Say

“By using SingularityPRO, we will be able to facilitate Fugaku to be a secure, reproducible and portable high-performance computing environment for a wide spectrum of users spanning academia to industry.”
Satoshi Matsuoka

Director of RIKEN Center for Computational Science (R-CCS)

“As an early adopter of Singularity, the San Diego Supercomputer Center (SDSC) has been collaborating with Sylabs and the development team for several years. The scientific community is developing software at an unprecedented rate, and our collaboration with Sylabs helps us more quickly respond to the evolving needs of our users.”
Mahidhar Tatineni

User Support Group Lead at San Diego Supercomputer Center

“The availability of ready-to-use binary distributions is the primary benefit we’ve derived from SingularityPRO. This has made it much easier for our systems configuration staff to provision SingularityPRO onto servers and compute nodes."
Oleksandr Moskalenko

Research Computing Applications and Facilitation Lead, University of Florida

“When we first adopted Singularity a few years ago, the driving case was to support TensorFlow on our cluster with GPUs and an old operating system. It was relatively easy to build a container with the right libraries and environment using Singularity. The use cases today are increasing even as we update our clusters, so Singularity is more than a transient solution.”
Chris Reidy

Research Facilitation Manager, University of Arizona

“Reproducibility is especially important to ensure scientific integrity and accuracy, making Singularity containers incredibly valuable because they can be shared so seamlessly. Along with many other bioinformatics researchers (too many to name), we are helping set a new standard for scientific discovery with our commitment to ensuring our studies are entirely reproducible with Singularity containers from Sylabs.”
Mark T. W. Ebbert, Ph.D

Assistant Professor Sanders-Brown Center, University of Kentucky

“As our advanced simulation codes push the boundaries towards more specialized hardware and systems, the need for performance portability and adaptability grows in HPC. It's about ensuring our software doesn't just run, but runs efficiently and consistently, regardless of the underlying system. SingularityCE 4.0 addresses this vision in an open-source context, offering interoperable OCI support and facilitating adaptable and efficient containerized application deployments. Amidst the complexities of large-scale computing environments, teams thrive when collaboration is seamless; SingularityCE 4.0 takes a commendable step in facilitating the power of teamwork."
Andrew J Younge

R&D Manager of the Scalable Computer Architectures, Sandia National Laboratories