Time Series forecasting for 20,000 products using python

I am using timeseries forecasting(ARIMA) to forecast the future demands of products of a store handleing 20,000 variety of products. Currently different models are developed and used to forecast future demands of different products.To predict the future sales of 20,000 products it is taking around 4 hours. This does not seems to be the best way of predicting future sales of 20,000 products.

Can a single model be developed which can handle all the 20,000 products?

Can you please help me by suggesting what should be my approach which handleing this large set of products and predicting their future sales?

I am new to this platform, kindly let me know in case of any mistakes.

Thanks,