Skill UI

A comprehensive framework for ensuring rigorous scientific reproducibility in systems papers. It details how to establish a full provenance ledger (tracking hardware, software commits, and workloads), manage performance variance through repeated runs and statistical reporting, and implement automated pipelines so that all reported figures can be regenerated from raw data.

A comprehensive audit checklist for submitting a paper to the HPCA conference. This guide covers critical checkpoints including the two-step abstract-and-conflict registration, adherence to the IEEE template (11-page limit), maintaining double-blind anonymity, ensuring full author listing in references, and disclosing AI-generated content. Use this to mitigate desk-reject risks and ensure timely submission readiness.

A comprehensive guide on achieving scientific reproducibility for complex computer architecture and system simulations. It details how to manage and pin all dependencies—including simulator commit hashes, local patches, workload binaries, hardware states, and post-processing scripts—to ensure that every reported figure and number can be accurately regenerated, regardless of time or environment changes.