Sustainable Computing: Energy Efficiency and Environmental Impact

The field of computing is shifting towards sustainable and energy-efficient solutions. Recent developments focus on optimizing energy consumption, reducing environmental impact, and promoting eco-friendly technologies. Researchers are exploring innovative approaches to minimize energy waste, such as dynamic job time limit adjustment and automated energy measurement. Additionally, there is a growing emphasis on quantifying the environmental footprint of computing systems, including the use of per- and poly-fluoroalkyl substances (PFAS) in semiconductor manufacturing. Noteworthy papers in this area include the ML.ENERGY Benchmark, which provides a framework for measuring inference energy consumption, and the study on benchmarking the energy, water, and carbon footprint of large language models. These advancements have the potential to significantly reduce the environmental impact of computing and pave the way for a more sustainable future.

Sources

Characterizing GPU Energy Usage in Exascale-Ready Portable Science Applications

An Autonomy Loop for Dynamic HPC Job Time Limit Adjustment

The ML.ENERGY Benchmark: Toward Automated Inference Energy Measurement and Optimization

Modeling PFAS in Semiconductor Manufacturing to Quantify Trade-offs in Energy Efficiency and Environmental Impact of Computing Systems

EcoSphere: A Decision-Support Tool for Automated Carbon Emission and Cost Optimization in Sustainable Urban Development

Strategies to Measure Energy Consumption Using RAPL During Workflow Execution on Commodity Clusters

How Hungry is AI? Benchmarking Energy, Water, and Carbon Footprint of LLM Inference

AI Greenferencing: Routing AI Inferencing to Green Modular Data Centers with Heron

Built with on top of