Search for: All records

We develop a comprehensive framework for storing, analyzing, forecasting, and visualizing industrial energy systems consisting of multiple devices and sensors. Our framework models complex energy systems as a dynamic knowledge graph, utilizes a novel machine learning (ML) model for energy forecasting, and visualizes continuous predictions through an interactive dashboard. At the core of this framework is A-RNN, a simple yet efficient model that uses dynamic attention mechanisms for automated feature selection. We validate the model using datasets from two manufacturers and one university testbed containing hundreds of sensors. Our results show that A-RNN forecasts energy usage within 5% of observed values. These enhanced predictions are as much as 50% more accurate than those produced by standard RNN models that rely on individual features and devices. Additionally, A-RNN identifies key features that impact forecasting accuracy, providing interpretability for model forecasts. Our analytics platform is computationally and memory efficient, making it suitable for deployment on edge devices and in manufacturing plants. 

The carbon footprint associated with large language models (LLMs) is a significant concern, encompassing emissions from their training, inference, experimentation, and storage processes, including operational and embodied carbon emissions. An essential aspect is accurately estimating the carbon impact of emerging LLMs even before their training, which heavily relies on GPU usage. Existing studies have reported the carbon footprint of LLM training, but only one tool, mlco2, can predict the carbon footprint of new neural networks prior to physical training. However, mlco2 has several serious limitations. It cannot extend its estimation to dense or mixture-of-experts (MoE) LLMs, disregards critical architectural parameters, focuses solely on GPUs, and cannot model embodied carbon footprints. Addressing these gaps, we introduce \textit{\carb}, an end-to-end carbon footprint projection model designed for both dense and MoE LLMs. Compared to mlco2, \carb~significantly enhances the accuracy of carbon footprint estimations for various LLMs. The source code is released at \url{https://github.com/SotaroKaneda/MLCarbon}. 

ABSTRACT Gas in the central regions of cool-core clusters and other massive haloes has a short cooling time (≲1 Gyr). Theoretical models predict that this gas is susceptible to multiphase condensation, in which cold gas is expected to condense out of the hot phase if the ratio of the thermal instability growth time-scale (tti) to the free-fall time (tff) is tti/tff ≲ 10. The turbulent mixing time tmix is another important time-scale: if tmix is short enough, the fluctuations are mixed before they can cool. In this study, we perform high-resolution (5122 × 768–10242 × 1536 resolution elements) hydrodynamic simulations of turbulence in a stratified medium, including radiative cooling of the gas. We explore the parameter space of tti/tff and tti/tmix relevant to galaxy and cluster haloes. We also study the effect of the steepness of the entropy profile, the strength of turbulent forcing and the nature of turbulent forcing (natural mixture versus compressive modes) on multiphase gas condensation. We find that larger values of tti/tff or tti/tmix generally imply stability against multiphase gas condensation, whereas larger density fluctuations (e.g. due to compressible turbulence) promote multiphase gas condensation. We propose a new criterion min (tti/min (tmix, tff)) ≲ c2 × exp (c1σs) for when the halo becomes multiphase, where σs denotes the amplitude of logarithmic density fluctuations and c1 ≃ 6, c2 ≃ 1.8 from an empirical fit to our results.