Scholarlink

As technological progress drives a rapid increase in data volume, recent attention has shifted toward heterogeneous computing systems, particularly with high-end personal computers leveraging advanced graphics processing units (GPUs). However, this growing data volume significantly extends GPU processing times, making parallel processing essential for enhanced efficiency. Our study explores image processing and demonstrates that CUDA (Compute Unified Device Architecture) optimizes heterogeneous GPU-based systems for this purpose. By enabling simultaneous processing of multiple images, CUDA harnesses GPU parallelism effectively. Our research results highlight substantial speed improvements achieved through the combined parallelism of GPUs and central processing units (CPUs) using CUDA technology.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Accelerating Image Processing on GPU with CUDA Technology

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Mekhriddin Rakhimov,
  • Shakhzod Javliev,
  • Rashid Nasimov

摘要

As technological progress drives a rapid increase in data volume, recent attention has shifted toward heterogeneous computing systems, particularly with high-end personal computers leveraging advanced graphics processing units (GPUs). However, this growing data volume significantly extends GPU processing times, making parallel processing essential for enhanced efficiency. Our study explores image processing and demonstrates that CUDA (Compute Unified Device Architecture) optimizes heterogeneous GPU-based systems for this purpose. By enabling simultaneous processing of multiple images, CUDA harnesses GPU parallelism effectively. Our research results highlight substantial speed improvements achieved through the combined parallelism of GPUs and central processing units (CPUs) using CUDA technology.