BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:® I/O Insights and Recommendations for All
DTSTART;VALUE=DATE-TIME:20241203T141000Z
DTEND;VALUE=DATE-TIME:20241203T143000Z
DTSTAMP;VALUE=DATE-TIME:20260310T181949Z
UID:indico-contribution-2424@events.chpc.ac.za
DESCRIPTION:Speakers: Jean Luca Bez (Lawrence Berkeley National Laboratory
 )\nThe complexity of the HPC I/O stack combined with gaps in the state-of-
 the-art profiling tools creates a barrier that does not help end-users and
  scientific application developers solve the I/O performance problems they
  encounter. To bridge this gap\, we introduce Dristhi\, a multi-source int
 eractive analysis framework that empowers users to visualize I/O traces\, 
 identify bottlenecks\, and gain a deeper understanding of application beha
 vior. By combining cross-layer analysis with heuristic-based automatic det
 ection\, Dristhi provides actionable insights and recommendations to overc
 ome common I/O performance bottlenecks. This talk will delve into the desi
 gn and capabilities of Dristhi\, demonstrate its use in pinpointing I/O pe
 rformance issues\, and highlight upcoming features that will further enhan
 ce its functionality.\n\nhttps://events.chpc.ac.za/event/139/contributions
 /2424/
LOCATION:Boardwalk Convention Centre BICC.G-D1 - D1 Tsitsikamma
URL:https://events.chpc.ac.za/event/139/contributions/2424/
END:VEVENT
END:VCALENDAR