Weather & Climate

The field of numerical weather forecasting and climate modeling is relied upon nearly as much as oil exploration, in terms of impact to the global economy and society. Advances in this area in recent decades have resulted in far more accurate forecasts.

However, as many centers approach petascale performance levels (and processor core counts, main memory sizes and I/O requirements continue to grow), significant operational challenges emerge. Without an HPC workload scheduler, the result is unreliable runtimes, sluggish or inconsistent application performance, cost implications around memory requirements, and risk in meeting forecast schedules and maintaining service levels.

Quote
“With PBS Professional, we have an enterprise-grade scheduler we can count on, without adding overhead to our IT staff. An open source scheduler just doesn’t provide the same value.”

– INGV View Case Study

Running HPC systems and clusters with Altair’s PBS Professional scheduler provides weather prediction centers with a rich solution for optimized forecasting. The solution ensures high-performance computing (HPC) resources are more efficiently scheduled to increase system utilization, improve application performance and throughput, and enhance quality of service.

Benefits include:

Improved Quality of Service for Operations:

• Greater availability of resources

• Reliable and repeatable runtimes

• Improved ability to forecast when a job will run

• Reduced execution times to meet production schedules

• Fast turnaround of high-priority jobs

• Support for long-range scheduling

Optimized Performance:

• Optimized application throughput and overall system efficiency

• Optimized resource availability

• Avoidance of system thrashing

Improved Manageability:

• Reduced need for process-level management

• Configurability to easily implement custom scheduling behaviors

Request Informaton



Get Started Today!