Research map / Graph learning / Graph self-supervised learning
PERFOGRAPH: A Numerical Aware Program Graph Representation for Performance Optimization and Program Analysis
Ali TehraniJamsaz, Quazi Ishtiaque Mahmud, Le Chen, Nesreen K. Ahmed, Ali Jannesari
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Add aggregate data type nodes (ablation) | Perfograph · Device Mapping AMD | error (%): 13.0 -> 6.0 | Improved |
| Add aggregate data type nodes (ablation) | Perfograph · Device Mapping NVIDIA | error (%): 15.0 -> 10.0 | Improved |
| Add digit embedding (ablation) | Perfograph · Device Mapping AMD | error (%): 7.1 -> 6.0 | Improved |
| Add digit embedding (ablation) | Perfograph · Device Mapping NVIDIA | error (%): 10.6 -> 10.0 | Improved |
| Use Perfograph instead of ProGraML | Device Mapping AMD · Device Mapping AMD | error rate: 7.4% | Improved |
| Use Perfograph instead of ProGraML | Parallelism Discovery · OMP_Serial Pluto and autoPar subsets | accuracy: 2% | Improved |
| Use Perfograph instead of ProGraML | Parallelism Discovery · OMP_Serial DiscoPoP subset | accuracy: 9% | Improved |
| Use Perfograph instead of ProGraML | Parallel Pattern Detection · OMP_Serial | accuracy: 3% | Improved |
| Use Perfograph instead of ProGraML | NUMA and Prefetchers Configuration Prediction · Sandy Bridge | error rate: 3.5% | Improved |
| Use Perfograph instead of ProGraML | NUMA and Prefetchers Configuration Prediction · Skylake | error rate: 1.8% | Improved |
| Use Perfograph instead of ProGraML | Thread Coarsening Factor Prediction · Four computing platforms | correct TCF predictions: 9 -> 17 | Improved |
| Use Perfograph instead of inst2vec | Thread Coarsening Factor Prediction · AMD Radeon HD 5900 | speedup: 1.37 -> 1.19 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- ProGraML — ProGraML: A Graph-based Program Representation for Data Flow Analysis and Compiler Optimizations (ICML 2021) · better on 5 of 6 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- kGym: A Platform and Dataset to Benchmark Large Language Models on Linux Kernel Crash Resolution (NeurIPS 2024 · similarity 0.82)
- LayerDAG: A Layerwise Autoregressive Diffusion Model for Directed Acyclic Graph Generation (ICLR 2025 · similarity 0.82)
- RepoGraph: Enhancing AI Software Engineering with Repository-level Code Graph (ICLR 2025 · similarity 0.81)
- SWE-fficiency: Can Language Models Optimize Real-World Repositories on Real Workloads? (ICML 2026 · similarity 0.81)
- A Structure-Aware Framework for Learning Device Placements on Computation Graphs (NeurIPS 2024 · similarity 0.81)
- Improving Parallel Program Performance with LLM Optimizers via Agent-System Interfaces (ICML 2025 · similarity 0.81)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
