BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Asia/Hong_Kong
X-LIC-LOCATION:Asia/Hong_Kong
BEGIN:STANDARD
TZOFFSETFROM:+0800
TZOFFSETTO:+0800
TZNAME:HKT
DTSTART:19911015T033000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20251218T030656Z
LOCATION:Meeting Room S426+S427\, Level 4
DTSTART;TZID=Asia/Hong_Kong:20251218T112300
DTEND;TZID=Asia/Hong_Kong:20251218T113400
UID:siggraphasia_SIGGRAPH Asia 2025_sess152_papers_2424@linklings.com
SUMMARY:TC-GS: A Faster Gaussian Splatting Module Utilizing Tensor Cores
DESCRIPTION:Zimu Liao (Shanghai Jiao Tong University, Shanghai Artificial 
 Intelligence Laboratory); Jifeng Ding and Siwei Cui (Shanghai Artificial I
 ntelligence Laboratory, Fudan University); Ruixuan Gong (Beijing Institute
  of Technology); Boni Hu (Shanghai Artificial Intelligence Laboratory, Nor
 thwestern Polytechnical University); Yi Wang (Shanghai Artificial Intellig
 ence Laboratory); Hengjie Li (Shanghai Artificial Intelligence Laboratory,
  Shanghai Innovation Insititute); and Hui Wang, Xingcheng Zhang, and Rong 
 Fu (Shanghai Artificial Intelligence Laboratory)\n\n3D Gaussian Splatting 
 (3DGS) renders pixels by rasterizing Gaussian primitives, where conditiona
 l alpha-blending dominates the time cost in the rendering pipeline. This p
 aper proposes TC-GS, an algorithm-independent universal module that expand
 s Tensor Core (TCU) applicability for 3DGS, leading to substantial speedup
 s and seamless integration into existing 3DGS optimization frameworks. The
  key innovation lies in mapping alpha computation to matrix multiplication
 , fully utilizing otherwise idle TCUs in existing 3DGS implementations. TC
 -GS provides plug-and-play acceleration for existing top-tier acceleration
  algorithms tightly coupled with rendering pipeline designs, like Gaussian
  compression and redundancy elimination algorithms. Additionally, we intro
 duce a global-to-local coordinate transformation to mitigate rounding erro
 rs from quadratic terms of pixel coordinates caused by Tensor Core half-pr
 ecision computation. Extensive experiments demonstrate that our method mai
 ntains rendering quality while providing an additional 2.18× speedup over 
 existing Gaussian acceleration algorithms, thus reaching up to a total 5.6
 × acceleration.\n\nRegistration Category: Full Access, Full Access Support
 er\n\nSession Chair: Lin Gao (University of Chinese Academy of Sciences)\n
 \n
END:VEVENT
END:VCALENDAR