Stardust: Scalable and Transferable Workload Mapping for Large AI on Multi-Chiplet Systems

Le résumé fourni par la source

Workload partitioning and mapping are critical to optimizing performance in multi-chiplet systems. However, existing approaches struggle with scalability in large search spaces and lack transferability across different workloads. To overcome these limitations, we proposeStardust, ascalable andtransferable workloadmapping on multi-chiplet systems.Stardustcombines learnable graph clustering to downscale computation graphs for efficient partitioning, topology-masked attention to capture structural information, and deep reinforcement learning (DRL) for optimized workload mapping. Evaluations on production-scale AI models show that (1)Stardust-generated mappings significantly outperform commonly used heuristics in throughput, and (2) fine-tuning a pre-trainedStardustmodel improves sample efficiency by up to 15× compared to training from scratch.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.