# A Distributed Framework for Causal Modeling of Performance Variability in GPU Traces

arXiv: https://arxiv.org/abs/2510.18300 · HTML: https://mlsanity.com/paper/2510.18300/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 6 of its 18 references that are in the map.
- First posted on arXiv 21 Oct 2025.

## Builds on (most cited first, 6 of 6)

- [Hierarchical Roofline Performance Analysis for Deep Learning Applications](https://mlsanity.com/paper/2009.05257/) (2020)
- [PGT-I: Scaling Spatiotemporal GNNs with Memory-Efficient Distributed Training](https://mlsanity.com/paper/2507.11683/) (2025)
- [KLARAPTOR: A Tool for Dynamically Finding Optimal Kernel Launch Parameters Targeting CUDA Programs](https://mlsanity.com/paper/1911.02373/) (2019)
- [Metrics and Design of an Instruction Roofline Model for AMD GPUs](https://mlsanity.com/paper/2110.08221/) (2021)
- [Understanding GPU Resource Interference One Level Deeper](https://mlsanity.com/paper/2501.16909/) (2025)
- [WANDER: An Explainable Decision-Support Framework for HPC](https://mlsanity.com/paper/2506.04049/) (2025)

## Abstract (arXiv)

Large-scale GPU traces play a critical role in identifying performance bottlenecks within heterogeneous High-Performance Computing (HPC) architectures. However, the sheer volume and complexity of a single trace of data make performance analysis both computationally expensive and time-consuming. To address this challenge, we present an end-to-end parallel performance analysis framework designed to handle multiple large-scale GPU traces efficiently. Our proposed framework partitions and processes trace data concurrently and employs causal graph methods and parallel coordinating chart to expose performance variability and dependencies across execution flows. Experimental results demonstrate a 67% improvement in terms of scalability, highlighting the effectiveness of our pipeline for analyzing multiple traces independently.