Sylabs introduces Singularity Pro
Commercial startup Sylabs recently launched an open source container technology called Singularity Pro made for high-performance computing (HPC). The Linux-based Singularity container platform serves as the basis of the technology, which was originally designed for HPC and scientific use cases.
Singularity started as an open source project in 2015 and is currently in its 13th release, version 2.4.2, running more than one million containers per day with the Open Science Grid. CEO Gregory Kurtzer, who created the Singularity, founded Sylabs last year. The company reports its user base exceeds 25,000, with users at academic institutions like Stanford University, Ohio State University and Michigan State, as well as at HPC centers like the Texas Advanced Computing Center, the San Diego Supercomputer Center, Oak Ridge National Lab and Lawrence Berkeley National Lab.
Containers are a lightweight form of virtualization, which divide a monolithic application into suite a modules. Early containers originally revolved around microservices. HPC, however, focuses on processing jobs, which require aggregating computing power to solve complex, computational problems. The technology is currently being applied to emerging fields like artificial intelligence (A.I.), deep learning and advanced analytics.
Among the key features provided by Singularity, according to the company, include: allowing users to run containers without the security implications of granting users control of a root-owned daemon process or kernel; a single file (SIF) that encapsulates the runtime environment; support for support GPUs, Infiniband and Intel Phi natively; and compatibility with Docker Hub, among other tools.
“Already the container platform of choice by academia and commercial HPC centers, Singularity’s features also make it the ideal container technology for artificial intelligence, machine / deep learning, compute-driven analytics, and data science — areas that we characterize as enterprise performance computing, or EPC,” said Kurtzer in a statement. “These applications carry data-intensive workloads that demand HPC-like resources, and, as more companies leverage data to support their businesses, the need to properly containerize and support those workflows has grown substantially.”
