BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//hacksw/handcal//NONSGML v1.0//EN
CALSCALE:GREGORIAN
BEGIN:VEVENT
DTEND:20200110T220000Z
UID: 69f2547857822
DTSTAMP:20260429T135656Z
LOCATION:Olin Center for Educational Technology
DESCRIPTION:Guest speaker&nbsp\;Omar Aaziz will present "Enabling HPC Performance Insights via Monitoring and Analysis."\n\nDescription: During high performance computing (HPC) system acquisition\, a significant consideration&nbsp\;is given to the desired performance\, which drives the selection of processing components\,&nbsp\;memory\, interconnects\, file systems\, and more. The achieved performance\, however\,&nbsp\;is highly dependent on operational conditions\, applications\, and workloads\, of which&nbsp\;the discovery and assessment are critical in practice.\n\nAs such\, HPC system analysis has&nbsp\;been a long-standing need for administrators to observe the health of their systems\,&nbsp\;detect abnormal conditions\, and to take informed actions when restoring system&nbsp\;health. Moreover\, users strive to understand how well their jobs run and what are the&nbsp\;architecture limits that restrict the performance of their jobs.\n\nIn my talk\, I will present two&nbsp\;research projects. I will introduce a methodology for modeling the expected runtime of an HPC job based on historical application data and data from the job itself using the Neural&nbsp\;Network technique. This estimation model is useful for both HPC users and administrators&nbsp\;as a metric to establish a measure of performance of the job.\n\nI will then present a&nbsp\;data-driven methodology to characterize the relationship between parent and proxy&nbsp\;applications based on collecting runtime data from both and then using the data analytics&nbsp\;to find their correspondence or divergence. I will explain how to measure the dynamic&nbsp\;hardware behaviors and relationships between several parent and proxy application pairs&nbsp\;and evaluate several different metrics over the time-varying behavior that are potentially&nbsp\;useful for comparing real applications to their proxy counterparts\n\n&nbsp\;\n
URL;VALUE=URI:https://augustana.edu/about-us/events/2020/enabling-high-performance-computing-insights-monitoring-and-analysis
SUMMARY:Enabling High Performance Computing Insights via Monitoring and Analysis
DTSTART:20200110T220000Z
END:VEVENT
END:VCALENDAR