Home / Products / Platform LSF Session Scheduler
 

Overview

 

Extreme performance complements simplicity of workload scheduling

 

Platform LSF Session Scheduler, a new add-on for Platform LSF, gives users the best of both worlds – speed and volume. Platform LSF Session Scheduler is a breakthrough innovation that brings a new low-latency, high-throughput scheduling paradigm to clustered computing environments. Both Platform LSF and Platform LSF License Scheduler are part of the new Platform Accelerate suite of products.

With the new session scheduler, customers can realize the benefits of high-throughput, low-latency task submission without needing to re-architect applications. Whether your large-scale computing problem is in semiconductor design analysis, vehicle crash simulation or pharmaceutical research, experienced high-performance datacenter managers will appreciate this important new capability.


 

How you can benefit from Platform LSF Session Scheduler:

 

  • Significantly improve overall cluster performance on clusters with a heavy short job load
  • Produce results faster than ever
  • Improve designs with more detailed simulations
  • Improve productivity
  • Reduce complexity in your HPC environment
  • Extend your investment in Platform LSF

 

Facts about Platform LSF Session Scheduler:

 

  • This new add-on is ideal for customers who deal with increasing numbers of short-duration jobs, have latency-bound workloads, or run large numbers of related simulation jobs
  • Provides a short-duration task engine with no code changes required, and no disruption to existing job semantics
  • Hundreds or even thousands of users can now process jobs individually comprised of over 50,000 tasks per user with exceptionally low-latency and minimal involvement of the Platform LSF workload scheduler.
  • Customers employing this scheduling approach can continue to use familiar Platform LSF features including pre- and post-execution scripts, job starters and job arrays.
  • By fully supporting Platform LSF job semantics, existing scripts can be easily modified to take advantage of this powerful new scheduling approach.