Predict

Accurately Visualize the Energy Road Ahead for Your Data Centers

Carbon Relay’s Predict product also helps customers boost their data center energy efficiency to reduce consumption and costs, as well as CO2 emissions. Predict gives data center operations managers and facilities professionals predictive intelligence about their environments’ energy consumption to guide them to the right operational decisions for maximizing the energy efficiency of their data centers.

Deployment is an easy and straightforward process with only minimal hardware integration required.  Predict leverages the historical data collected by the thousands of IoT sensors inside data centers to train itself. Predict uses this data–such as temperatures, power levels, and airflow rates – to orient its models on power usage effectiveness (PUE), and then to predict sensor readings for the immediate future (the next hour, day or week).

Through integration with data visualization tools, Predict highlights efficiency opportunities via dashboards with intuitive graphics.  This makes it easy for managers to focus on and prioritize these items and take the appropriate actions. Predict’s insights can also help data center managers intervene early with energy-related operational issues and take swift action to prevent them from becoming larger problems.

Laptop
Unlike our Optimize solution, Predict requires customers to provide historical operations data upfront to train the predictive module. On the other hand, deployment is an easy and straightforward process with only minimal hardware integration required. That’s because Predict can leverage the historical data – such as temperatures, power levels, and flow rates –collected by the thousands of IoT sensors inside data centers. It uses that data to orient its models on power usage effectiveness (PUE) and trains its models to predict sensor readings for the immediate future.

Predict’s recommendations guide human operators to the right operational decisions for maximizing the energy efficiency of their data centers.

Once Predict customers are up and running, they gain the ability to accurately forecast the performance of any of the hundreds of variables it monitors.  Whether it’s for the next two hours or the next two days, managers have the energy-centric visibility and insights they need to inform their decision making.  
1 standard deviation of the prediction is displayed in pink and the actual data is in cyan. Displayed is the real output of our model predicting PUE.

Your next step

Want to see what our products can do in your environment? It’s a no-cost, no-risk way for you to see first-hand the energy efficiency gains our products will create for you. We’d be happy to discuss further.

Just complete this form, and we will be in touch to get it scheduled.