Scholarlink

The growing disparity between system processing power and I/O and storage capabilities has underscored the importance of in situ visualization and analysis. However, adoption of these technologies remains limited due to hesitancy around new coding techniques, perceived workflow complexity, and concerns over resource costs. To address this and better educate users, we previously introduced Inshimtu, an in situ “shim” library designed to facilitate experimentation with in situ techniques and serve as a pedagogical tool, allowing users to experience in situ workflows before committing to full-scale integration. Building on this work, we have updated Inshimtu to utilize the Catalyst2 standard. We have also developed a version of the Gray-Scott reaction-diffusion MiniApp—a two-variable application using a memory-bound 7-point stencil kernel—to demonstrate in situ workflows at scale. This paper details the process of converting Inshimtu from Legacy ParaView Catalyst to Catalyst2, discusses the complexities and lessons learned, describes the integration of Inshimtu with the Gray-Scott MiniApp for scalability testing, and demonstrates the combined functionality within a scientific workflow on the Shaheen III supercomputer. Inshimtu is available for download at: https://github.com/kaust-vislab/Inshimtu-basic Our version of the Gray-Scott MiniApp is available for download at: https://gitlab.kitware.com/jameskress/KAUST_Visualization_Vignettes .

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

Updating Inshimtu with Catalyst2 and Integrating an HPC MiniApp: Lessons Learned

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • James Kress,
  • Thomas Theußl

摘要

The growing disparity between system processing power and I/O and storage capabilities has underscored the importance of in situ visualization and analysis. However, adoption of these technologies remains limited due to hesitancy around new coding techniques, perceived workflow complexity, and concerns over resource costs. To address this and better educate users, we previously introduced Inshimtu, an in situ “shim” library designed to facilitate experimentation with in situ techniques and serve as a pedagogical tool, allowing users to experience in situ workflows before committing to full-scale integration. Building on this work, we have updated Inshimtu to utilize the Catalyst2 standard. We have also developed a version of the Gray-Scott reaction-diffusion MiniApp—a two-variable application using a memory-bound 7-point stencil kernel—to demonstrate in situ workflows at scale. This paper details the process of converting Inshimtu from Legacy ParaView Catalyst to Catalyst2, discusses the complexities and lessons learned, describes the integration of Inshimtu with the Gray-Scott MiniApp for scalability testing, and demonstrates the combined functionality within a scientific workflow on the Shaheen III supercomputer. Inshimtu is available for download at: https://github.com/kaust-vislab/Inshimtu-basic Our version of the Gray-Scott MiniApp is available for download at: https://gitlab.kitware.com/jameskress/KAUST_Visualization_Vignettes .