PulseTrendingJournal ClubResearchersJournalsExplore
Instagram
HomeTrendingJournal ClubExplore
Synapse
⌘+K
Synapse
April 8, 2026ComputingOpen Access

Adaptive mixed-precision Monte Carlo integration on GPUs

View Full Paper
Ask AI
Bookmark
Share

Authors

FÖFerhat Onur ÖzganBKBerke KabasakalFTF. Sukru Torun

Discussion

Loading...

Member takes

Overview

This framework enhances GPU-accelerated Monte Carlo integration accuracy and performance by adaptively selecting precision levels.

Key Points

  • The aim is to enhance Monte Carlo integration efficiency on GPUs by adaptively selecting precision for function evaluations.
  • Developed a GPU-accelerated mixed-precision MCI framework.
  • Implemented adaptive precision selection based on local numerical characteristics.
  • Explored term-wise and region-wise precision allocation strategies using CUDA batch processing.
  • Achieved speedups of up to 4.9 times compared to full double precision.
  • Maintained controlled relative error across varying dimensions of test functions.

Cite This Study

Özgan et al. (2026) studied this question.

synapsesocial.com/papers/69d5f0d774eaea4b11a7a51fhttps://doi.org/10.1007/s00607-026-01656-7
View Full Paper
Ask AI
Bookmark
Share