- taskvine
- sade
- floability
- pledge
- xgfabric
- highlights
•
•
•
•
•
-
Introducing X-Bucket
In cluster-backed dynamic workflows, the allocator must reserve cores, memory, and related resources before each task runs, but its real footprint only shows up afterward. Dynamic workflows make that puzzle even messier. X-Bucket turns streaming measurements into bucketed predictions that stay general, free of priors, and online. Read on for how it stacks up.
-
TaskVine Insights: How to Propose a Pull Request to the CCL Team
Contributions help TaskVine and CCTools move forward. This post welcomes discussion, pull requests, and patches, then describes the workflow from issue to merge and how to build, test, and submit against the official repository.
-
Cursor Rules in Research Computing
One of our lab researchers shares practical experience with Cursor rules and AGENTS.md-style project context: how they help AI assistants stay aligned with style, layout, and how we run tests in research computing workflows.
-
TaskVine Insights: A Beginner's Map to the CCTools Codebase and a First Task Dispatch Patch
CCTools can look intimidating at first glance. This post gives beginners a practical map of the repository, explains how TaskVine fits into the larger system, and walks through a small but real source change that prints which task was committed to which worker.
-
CCL Participated in the 18th ND CSE Annual Poster Session
CCL team members presented eight research posters at the 18th ND CSE Annual Poster Session on February 26, 2026, spanning HEP workflows, DAG execution, reproducibility, digital agriculture, and LLM inference.