Research map / Optimization / Nonconvex and smooth optimization
Delay-agnostic Asynchronous Coordinate Update Algorithm
Xuyang Wu, Changxin Liu, Sindri Magnússon, Mikael Johansson
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 |
|---|---|---|---|
| Replace ARock's delay-dependent step-size with DEGAS's delay-free parameters | asynchronous BCD and ADMM · Lasso and logistic regression on CIFAR100 | convergence speed in # of Ti computed | Improved |
| Use small-delay-dominated delay distribution instead of large-delay distribution (ablation) | DEGAS on simple operator T(x)=0.8x · simulated delays (m=20, max delay 20) | convergence rate of iterates | Improved |
| Synchronous update instead of asynchronous DEGAS | coordinate update in Lasso and logistic regression · CIFAR100 experiments | convergence per # of Ti computed | Mixed |
| Tune ARock step-size in its theoretical range (vs. DEGAS with no tuning) (ablation) | ARock · simulated operator and CIFAR100 experiments | convergence speed | No clear effect |
Closest papers by meaning
- Asynchronous Coordinate Descent under More Realistic Assumptions (NeurIPS 2017 · similarity 0.89)
- Asynchronous Iterations in Optimization: New Sequence Results and Sharper Algorithmic Guarantees (JMLR 2023 · similarity 0.87)
- A2BCD: Asynchronous Acceleration with Optimal Complexity (ICLR 2019 · similarity 0.86)
- Distributed Asynchronous Optimization with Unbounded Delays: How Slow Can You Go? (ICML 2018 · similarity 0.86)
- Faster Stochastic Optimization with Arbitrary Delays via Adaptive Asynchronous Mini-Batching (ICML 2025 · similarity 0.85)
- Do We Need Asynchronous SGD? On the Near-Optimality of Synchronous Methods (NeurIPS 2026 · similarity 0.84)
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.
