Architected to optimize throughput for all computing infrastructure; WorkloadXelerator™ works as an upper level scheduler for NetworkComputer™, IBM® Platform™ LSF® and other third party schedulers
SANTA CLARA, Calif. — (BUSINESS WIRE) — June 6, 2016 — Runtime, a leading provider of infrastructure and design optimization software solutions, introduced a new flexible hierarchical scheduler, WorkloadXelerator. WorkloadXelerator is a high-performance hierarchical scheduler designed for distributed high-performance computing (HPC) environments.
Today’s system on chip (SoC) design organizations manage a variety of job mixes including several long jobs for physical design and millions of short jobs for design verification and library characterization. Many job schedulers cannot support this number of short jobs as their queue size is limited. In addition, the dispatch rate of many base schedulers is too low due to their cycle-based architecture, which limits them to dispatching jobs at regular intervals. WorkloadXelerator has millisecond dispatch latency for job scheduling and is designed specifically to manage EDA job mixes by providing one or more upper level “personal schedulers.” These are managed directly by individual design or project teams to run flows for design verification, IPQA, and other demanding workloads.
“We continue to work closely with our customers as they scale up design verification activities,” said Dr. Andrea Casotto, founder and CEO of Runtime. “Our family of high-performance schedulers are uniquely positioned to maximize throughput and capacity utilization. For teams with difficult workloads, WorkloadXelerator allows full access to all available compute resources without artificial constraints imposed by limits of the base scheduler, thus speeding up the design process and maximizing utilization of resources.”
WorkloadXelerator incorporates the same event-driven core technology as NetworkComputer and is designed to reduce the base scheduler’s job submission and creation, job dispatch, user queries, and job queue size. It is also equipped with a comprehensive set of policy management features including FairShare and preemption. Multiple WorkloadXelerators can be used in conjunction with a base scheduler to further increase the queue size. Typical deployments of WorkloadXelerator realize a 6-10X increase in throughput with no added compute capacity over the existing base scheduler.
WorkloadXelerator is available now. Runtime is demonstrating WorkloadXelerator at the 53rd Design Automation Conference (DAC) at the Austin Convention Center in Austin, Texas from June 6-8, 2016 in Booth #1728. Please contact your Runtime sales representative to schedule a private appointment.
Runtime, headquartered in Santa Clara, California, is the leading supplier of high-performance computing infrastructure and workflow optimization solutions, dramatically improving resource utilization and efficiency and enabling faster time to market for our customers. We offer a unique product family specializing in high-performance job scheduling, workflow development, monitoring and execution, and efficient resource management. Our commitment to our diverse customer base, including semiconductor, mobile, IoT, and automotive suppliers, enables them to rapidly deliver leading-edge products that transform the world we live in today. For more information, visit www.rtda.com.
All trademarks are property of their respective owners.
Tiffany Quan, 650-550-0803