Support Vector Machines - Why is SVM not popular?

Support Vector Machines (SVMs) aren't the best option for dealing with massive datasets because of all the computer power they require. Datasets with a lot of noise or overlapping classifications might also be problematic for SVM because it depends on recognizing a clear separation margin between classes.