BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//pretalx.com//pydata-london-2026//talk//WGJMXV
BEGIN:VTIMEZONE
TZID:GMT
BEGIN:STANDARD
DTSTART:20001029T030000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10
TZNAME:GMT
TZOFFSETFROM:+0100
TZOFFSETTO:+0000
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20000326T020000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3
TZNAME:BST
TZOFFSETFROM:+0000
TZOFFSETTO:+0100
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
UID:pretalx-pydata-london-2026-WGJMXV@pretalx.com
DTSTART;TZID=GMT:20260606T110500
DTEND;TZID=GMT:20260606T115000
DESCRIPTION:When building high-performance systems for analytical workload\
 , we often focus on the efficiency of the algorithm\, like reducing Big-O 
 complexity or optimising numerical routines. Yet in real world workloads\,
  the decisive factor is not just the algorithm but the shape of how the da
 ta is laid out\, traversed\, and distributed across processes.\n \nThis ta
 lk will cover aspects of mechanical sympathy\, focussing on how structures
  in memory can benefit from cache-sensitive\, SIMD-enabled (vector instruc
 tions) CPUs\, constrained by memory bandwidth and optimised for predictabl
 e\, contiguous access.\n\nWe will use real-world examples to show how mini
 mising serialisation overhead and enabling efficient cross-process and cro
 ss-language data exchange reduces the cost of data movement across systems
 . Beyond single-system performance\, we will examine why Arrow’s standar
 dised\, zero-copy columnar format is a critical enabler of distributed exe
 cution. We will see how columnar formats support scalable computation acro
 ss threads\, processes\, and distributed nodes.
DTSTAMP:20260602T223343Z
LOCATION:Grand Hall 1
SUMMARY:Columnar Thinking - Designing for high-performance execution with A
 rrow and Polars - Kamlesh Shah
URL:https://pretalx.com/pydata-london-2026/talk/WGJMXV/
END:VEVENT
END:VCALENDAR