Note: This article originally appeared on InsideHPC, click here to read the original piece.
Today GigaIO announced availability of FabreX for vScaler’s management interface. As the industry’s first in-memory network, FabreX will bolster vScaler’s cloud appliances for artificial intelligence (AI), deep learning, biotechnology and big data analytics.
GigaIO entered a strategic partnership with vScaler in November 2019 to bring their excellent user interface and ease of use into the FabreX environment,” says Alan Benjamin, CEO of GigaIO. “FabreX’s integration with vScaler delivers an elegant and straight forward way for customers to improve resource utilization and create highly-composable, unified infrastructures. I am thrilled this optimization has finally come to fruition and is available to the general public.”
vScaler is a private cloud platform built on open source technology (OpenStack) that enables the design of a secure, scalable, cost effective and flexible infrastructure for high-performance computing (HPC) workloads of all sizes. The technology allows users to dynamically provision physical servers on demand from pools of bare-metal resources, and features application specific software stacks and frameworks that are capable of meeting highly specific size requirements.
The combination of FabreX and vScaler offers a flexible alternative paradigm where we can compose resources that efficiently map to applications, rather than providing static resources that are only partially-used,” comments David Power, director and chief technology officer of vScaler. “It is this dynamic configurability that makes FabreX an ideal partner solution for our cloud platform.”
FabreX enhances vScaler’s ability to enable autonomous high-performance computing workloads, delivering improved resource utilization and decreased latency in the computing cluster to minimize overall system costs. The switch integrates compute, acceleration, storage and input/output (IO) communication into a single‐system cluster to create the industry’s first in-memory network. FabreX facilitates direct communication between multiple servers, natively running Message Passing Interface (MPI) applications, containers and hypervisors so application servers can utilize accelerators and storage located in other servers without going through their respective central processing unit.
Additionally, FabreX for vScaler integrates with Slurm, a scalable, open-source management system for Linux clusters of all sizes. Built on OpenStack software, the optimization enables users to experience the benefits of system efficiency and agility that containerization affords, including access and support for containerized environments such as Docker and Kubernetes. Users will benefit from the feature rich environment and performance as they provision their workloads provided by FabreX and vScaler.
FabreX for vScaler is available now.