Choosing Hardware for Deep Learning Applications

I am an undergraduate researcher at Georgia Southern University. Recently I became involved in the field of deep learning through a research project that applied transfer learning to medical diagnosis.

Currently my lab utilizes a RTX 2080 GPU for all of our training needs. That said, recently we have obtained a grant that allows the purchase of deep learning hardware in the budget range of 15,000 to 20,000 dollars.

As I am relatively new to the field, I am looking for advice on the best platform to purchase for the price. I have researched some information on the Lambda Blade and Quadro systems, but honestly I have been inundated in all the possibilities. Future research to be done on this system will be varied, so it is not targeted at a specific application. Any input is appreciated.