BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//pretalx.com//pydata-london-2026//talk//ABYV3J
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-ABYV3J@pretalx.com
DTSTART;TZID=GMT:20260606T153000
DTEND;TZID=GMT:20260606T161500
DESCRIPTION:Choosing a cloud instance type for a DS/ML/AI workload is still
  largely a heuristic exercise. While public pricing and hardware specifica
 tions are available\, they are fragmented\, inconsistently structured\, an
 d challenging to compare across cloud providers -- especially once real wo
 rkload performance is taken into account.\n\nIn this talk\, we present Spa
 re Cores Navigator\, a Python-queryable benchmark dataset that covers thou
 sands of cloud server types from multiple vendors\, with standardized perf
 ormance and cost-efficiency metrics. We demonstrate how instance selection
  can be expressed as a simple data query\, e.g. filtering by workload char
 acteristics\, hardware or compliance constraints\, and budget\, then ranki
 ng candidates by price-performance.
DTSTAMP:20260602T223218Z
LOCATION:Grand Hall 1
SUMMARY:SELECT instance FROM cloud WHERE workload = ? ORDER BY cost_efficie
 ncy - Gergely Daroczi
URL:https://pretalx.com/pydata-london-2026/talk/ABYV3J/
END:VEVENT
END:VCALENDAR