How do we determine high dimensionality?

I understand that "high dimensionality" is likely relative to the number of features compared to the number of rows in our dataset but is there any rule of thumb for determining the upper limit of dimensionality when deciding whether or not to use SMOTE? Thank you.