arXiv preprint ·
A principled approach for energy-efficient training via phase-aware GPU frequency tuning
- Published
- Source
- arXiv preprint, posted before peer review
- Original
- Read it at arXiv
- Topics
Abstract
Modern AI model training imposes unprecedented computational demands, making it a key contributor to datacenter energy consumption. Yet a significant fraction of the energy consumed during training does not translate to useful computation due to bottlenecks throughout the training pipeline. We present PAFT, a phase-aware, dynamically adaptable GPU frequency tuning system that reduces energy consumption of training workloads with minimal performance overhead. The key insight behind PAFT is that bottlenecks represent an energy optimization opportunity, rather than purely a performance problem: when GPUs are bound to stall, PAFT opportunistically reduces their clock frequencies to match the pace of bottlenecked devices, saving energy without impacting execution time. PAFT achieves this by continuously monitoring pipeline behavior and applying fine-grained frequency adjustments, adapting to workload and system changes. Experiments conducted on twelve widely used models show that PAFT consistently outperforms all baselines, achieving energy savings of up to 46% with an average overhead of 4%.
The authors’ abstract, as arXiv lists it. arXiv’s listings, abstracts included, are free for anyone to reuse.
Read it on ReadSearch
Each opens after you create an account, and brings you back to this paper.
More in Distributed & Parallel Computing
Byzantine Causal Reliable Broadcast (BCRB) with Constant-Size Message Metadata
Quantum Advantage for Distributed Symmetry Breaking
MCP-GRANITE Benchmark: GRANularity Interface TEsting for MCP-Based LLM Agents
EMA: Elastic and Performance Transparent Memory Across GPUs
Crossflow: Prefill-Decode Elasticity for Agentic LLM Serving
Get the few papers worth your time in Distributed & Parallel Computing every morning.
ReadSearch reads the new papers in your field each day and brings you the few worth reading, each with its key finding in a sentence.