In recent years, energy usage in LED Video Wall Display Panels (LED-VWDPs) has increased massively; Predicting energy consumption is a challenging and crucial task for LED-VWDPs. Hence Real-time energy usage issues can be resolved by predicting future energy demand. Deep learning plays an important role in more accurate prediction in energy forecasting. In this article, two approaches are presented: the first makes use of a recurrent neural network (RNN), and the other utilizes a long short-term memory (LSTM) network.In comparison to other existing machine learning techniques, such as ARIMA and Facebook Prophet, Long Short-Term Memory (LSTM) in deep learning is better at handling time-series datasets and projecting future energy demand. It predicts the actual energy usage of LED-VWDP and forecasts the futureenergydemandofLED-VWDP. A vast dataset of LED-VWDP energy consumption is utilized in this paper. Through the proposed RNN and LSTM methods, users can identify the individual energy usage of LED-VWDP and predict its future energy demand.The results of the proposed methods are evaluated alongside those of the existing methods in order to forecast energy usage. The results are used to evaluate the performance of forecasting future energy demands, depending on the number of epochs. The accuracy of RNN and LSTM ranges from 82.02–95.86%. The predictions have been made for a period of two months, encompassing short-and mid-term forecasts.In evaluating the comparison of various machine and deep learning models, LSTM is found to be accurate with an average root mean square error of 0.5 in forecasting energy consumption.