There continues to be an increasing trend within IT and research organizations to migrate workloads to the Cloud in order to accommodate the demands of today's high-performance applications. With this in mind, Univa's innovative workload management and optimization solutions, Navops Launch and Grid Engine, are uniquely suited for industries such as life sciences, manufacturing, and AI/analytics that require the highest level of performance to support data-intensive and performance-hungry applications. Accelerating machine learning projects through the elimination of storage bottlenecks is of high interest to enterprises, and this collaboration will allow WekaIO's high-speed scale-out file storage technology to support computational needs that help users advance machine learning projects into production.
"Our flash-native, parallel file system scales to deliver all of the required bandwidth to the most demanding applications", stated Liran Zvibel, co-founder and CEO at WekaIO. "Univa has deep expertise as an HPC Cloud leader that interacts with hundreds of companies, and we look forward to helping them migrate their customers' HPC workloads to the cloud. In tandem with Univa, we will offer life sciences research organisations and enterprise users highly-accelerated storage capabilities that can help eliminate the concern for performance bottlenecks as they look to radically increase their computing power and bandwidth within their HPC and ML environments."
Univa also announced the release of Navops Launch 1.0, which offers enterprise customers more significant advancements in the migration of HPC workloads to the Cloud. This latest and most powerful hybrid HPC Cloud management product will help users simplify the extension of on-premise clusters into hybrid Cloud services to efficiently mesh public and private Cloud resources to address increasing workload demands, while delivering increased operational efficiencies. A unique and key component of this new version is that it uses automation applets that monitor and control cloud scaling and workload placement in hybrid Clouds. Navops Launch automation tightly integrates with Univa Grid Engine and its workload status and metrics collection to respond to changes in workload demand and cluster utilization by expanding or contracting the Cloud footprint in a dynamic fashion. Once configured, administrators have full control over tying cloud resource usage to a corresponding budget without requiring manual intervention.
In addition, Navops Launch's unique automation applets can integrate third-party data sources, including storage fabric like WekaIO's flash-native, parallel file system, as well as management systems, cluster and Cloud environment attributes. The use of these attributes can be tied together with application metrics via a series of automation applets to drive insight into the scaling, reaping, data movement, cost decisions and workload placement policies. Crucial aspects of data locality and data movement can be automated with an applet as well and tied to the requirements of workloads and storage architectures.
"It has been a banner year for Univa in helping our enterprise users migrate their HPC workloads to the Cloud", stated Gary Tyreman, president and CEO of Univa. "The Univa team has been working closely with our customers from 11 of the top 12 life science organisations, 22 major airlines, the top two oil and gas companies and dozens of educational institutions, to help optimize performance and manage their mission-critical applications. As we recognize the need for a high-performance scale-out file system, we are delighted to partner with WekaIO to integrate one of the world's fastest parallel file systems into our Navops Launch to give end customers the ability to maximize the throughput of their HPC environment and make it easier to move high-performance applications to the Cloud."