Research map / Architectures and efficiency / RNNs and normalization
Efficient Use of Limited-Memory Accelerators for Linear Learning on Heterogeneous Systems
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| use duality-gap-based selection instead of random selection | Ridge regression · epsilon | convergence improvement: 10× | Helped |
| use DuHL instead of the reference schemes | Lasso · 30GB Dogs vs. Cats dataset | convergence speed: 10× faster | Helped |
| use DuHL instead of the reference schemes | SVM · 30GB Dogs vs. Cats dataset | convergence speed: 10× faster | Helped |
| use DuHL instead of the reference schemes | Ridge regression · 30GB Dogs vs. Cats dataset | convergence | Helped |
| use stale gap-memory values rather than oracle duality gaps (ablation) | DuHL · epsilon, Lasso | rounds to reach suboptimality 10^{-4}: within a factor of two | Hurt |
| decrease the frequency of gap-memory updates (ablation) | DuHL · epsilon, Lasso | convergence speed | Hurt |
| use gap-based selection rather than uniform random selection | DuHL · 30GB Dogs vs. Cats dataset, Lasso and SVM | I/O operations | Helped |
| increase the freshness of gap-memory entries (ablation) | DuHL · epsilon, Lasso | data-column swaps | Helped |
Closest papers by meaning
- SpeedLoader: An I/O efficient scheme for heterogeneous and distributed LLM operation (NeurIPS 2024 · similarity 0.87)
- Cost-Efficient LLM Training with Lifetime-Aware Tensor Offloading via GPUDirect Storage (NeurIPS 2025 · similarity 0.86)
- Harnessing Manycore Processors with Distributed Memory for Accelerated Training of Sparse and Recurrent Models (AAAI 2024 · similarity 0.86)
- Poplar: Efficient Scaling of Distributed DNN Training on Heterogeneous GPU Clusters (AAAI 2025 · similarity 0.86)
- Pfeife: Automatic Pipeline Parallelism for PyTorch (ICML 2025 · similarity 0.86)
- Practical Offloading for Fine-Tuning LLM on Commodity GPU via Learned Sparse Projectors (AAAI 2025 · similarity 0.85)
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.