Home > Servers > Rack and Tower Servers > AMD > White Papers > Finish machine learning preparation tasks on Kubernetes containers in less time with the PowerEdge R7525 > Image preparation for machine learning: Process images in less time
Before you can feed data into a machine learning algorithm, you must first prepare the data. For image-based machine learning workloads, this preparation step commonly includes several chained image-processing operations such as scaling, rotation, conversions, and more. Sorting millions of images is standard fare for training algorithms for quality control in manufacturing, self-driving cars, medical diagnoses, and more—so, it pays to have a solution that can make these preparations as quickly as possible. The sooner your machine completes preparation, the sooner you can get to the next phase. Because machine learning is a cycle, a solution that saves time in preparation will save time on every iteration.
Figure 1 shows the time results from our image processing test. The Dell EMC PowerEdge R7525 required 55.8 percent less time to process 3.3 million images, taking 11 minutes, 12 seconds compared to 25 minutes, 22 seconds for the HPE ProLiant DL380 Gen10.