BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Asia/Seoul
X-LIC-LOCATION:Asia/Seoul
BEGIN:STANDARD
TZOFFSETFROM:+0900
TZOFFSETTO:+0900
TZNAME:KST
DTSTART:18871231T000000
DTSTART:19881009T020000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20230103T035311Z
LOCATION:Room 325-AB\, Level 3\, West Wing
DTSTART;TZID=Asia/Seoul:20221208T140000
DTEND;TZID=Asia/Seoul:20221208T153000
UID:siggraphasia_SIGGRAPH Asia 2022_sess169_papers_263@linklings.com
SUMMARY:LuisaRender: A High-Performance Rendering Framework with Layered a
 nd Unified Interfaces on Stream Architectures
DESCRIPTION:Technical Communications, Technical Papers\n\nLuisaRender: A H
 igh-Performance Rendering Framework with Layered and Unified Interfaces on
  Stream Architectures\n\nZheng, Zhou, Chen, Yan, Zhang...\n\nThe advanceme
 nts in hardware have drawn more attention than ever to high-quality offlin
 e rendering with modern stream processors, both in the industry and in res
 earch fields. However, the graphics APIs are fragmented and existing shadi
 ng languages lack high-level constructs such as polymorphism, which adds c
 omplexity to developing and maintaining cross-platform high-performance re
 nderers. We present LuisaRender, a high-performance rendering framework fo
 r modern stream-architecture hardware. Our main contribution is an express
 ive C++-embedded DSL for kernel programming with JIT code generation and c
 ompilation. We also implement a unified runtime layer with resource wrappe
 rs and an optimized Monte Carlo renderer. Experiments on test scenes show 
 that LuisaRender achieves much higher performance than existing research r
 enderers on modern graphics hardware, e.g., 5–11× faster than PBRT-v4 and 
 4–16× faster than Mitsuba 3.\n\nRegistration Category: FULL ACCESS, ON-DEM
 AND ACCESS\n\nLanguage: ENGLISH\n\nFormat: IN-PERSON, ON-DEMAND
URL:https://sa2022.siggraph.org/en/full-program/?id=papers_263&sess=sess16
 9
END:VEVENT
END:VCALENDAR