Breadcrumb

NSF funded the research for machine learning based thermal monitoring and run-time thermal management for manicure processors and chiplet designs

NSF funded the research of VSCLAB for for machine learning based thermal monitoring and run-time thermal management for manycore processors and chiplet designs. Prof. Sheldon Tan is the single PI for this research work with 500K budget for three years. 

Today’s high-performance processors, and even emerging mobile platforms, are more thermally constrained than ever before due to continuing increase in on-chip power densities. Emerging Chiplet-based heterogeneous integration further exacerbates the thermal problems as heat dissipation is limited due to stacking integration. 

This project will explore a new generation of data-driven real-time thermal monitoring and smart run-time thermal/power and reliability management techniques by harnessing the latest advances in machine leaning and numerical methods for commercial multi/many core processors and chiplet design. 

This award is  the culmination of a few year research efforts at VSCLAB, which pioneered a few novel machine learning based approaches to full chip thermal analysis methods (DATE19, ASPDAC20, ICCAD20, TCAD20, TC21), first-principle based full-chip power map estimation (DATE20, TCAD21),  deep reinforcement learning based real-time thermal/reliability management (MLCAD20), the graph convolution networks based thermal analysis for chiplet designs (ASPDAC22). 

See  https://www.nsf.gov/awardsearch/showAward?AWD_ID=2113928&HistoricalAwards=false for details.