BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//pretalx.devconf.info//devconf-us-2026//talk//EXRZ3P
BEGIN:VTIMEZONE
TZID:EST
BEGIN:STANDARD
DTSTART:20001029T030000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10;UNTIL=20061029T070000Z
TZNAME:EST
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
END:STANDARD
BEGIN:STANDARD
DTSTART:20071104T030000
RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=11
TZNAME:EST
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20000402T030000
RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=4;UNTIL=20060402T080000Z
TZNAME:EDT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
END:DAYLIGHT
BEGIN:DAYLIGHT
DTSTART:20070311T030000
RRULE:FREQ=YEARLY;BYDAY=2SU;BYMONTH=3
TZNAME:EDT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
UID:pretalx-devconf-us-2026-EXRZ3P@pretalx.devconf.info
DTSTART;TZID=EST:20260924T164000
DTEND;TZID=EST:20260924T171500
DESCRIPTION:Large Language Models are increasingly constrained by inference
  architecture rather than training. As workloads diversify\, traditional m
 onolithic serving stacks struggle with mixed latency targets\, long contex
 t prompts\, GPU memory pressure\, and inefficient batching. This talk pres
 ents two emerging patterns treating LLM inference as a distributed systems
  challenge.\n\n**Dynamic Prefill Decode (PD)** separates the compute-inten
 sive prefill phase from the memory-bound decode phase\, scaling each acros
 s specialized GPU pools to improve utilization and reduce tail latency.\n*
 *Semantic Routing** replaces static load balancing with inference-aware de
 cisions\, selecting the right model and infrastructure tier per request.\n
 \nWe explore tools like vLLM\, Ray Serve\, KServe\, and TensorRT-LLM.\nThe
  session includes a live demo showing semantic routing\, model selection\,
  latency differences\, and cost tradeoffs across multiple models.
DTSTAMP:20260727T174706Z
LOCATION:Ladd Room (Capacity 170)
SUMMARY:Emerging LLM Serving Architectures: Dynamic PD Disaggregation & Sem
 antic Routing - Ayushi Tiwari
URL:https://pretalx.devconf.info/devconf-us-2026/talk/EXRZ3P/
END:VEVENT
END:VCALENDAR