Why not make a small project before starting the big one. A simple blinking light could already be enough to understand the basics of pcb design (invest a few weeks up front to learn the basics and your main project will go much more smoothly.)
For bga design i suggest you read manufacturer application notes about bga, research papers regarding reflow soldering in general and bga in particular, industry standards, your manufacturers guidelines and capabilities (compare multiple ones), …
Researching stuff on your own is after all the main thing you should take with you from university. This is what really differentiates a graduate of a university from somebody finishing a trade school.
And don’t be afraid to ask professors of your university. You will never again be able to get support by such highly qualified people for free. (Just ask one that could have knowledge in that field they at least can tell you who really knows about this stuff. If what you do is part of a course or thesis start with the people who are responsible for it. If it is for a student team then ask alumni of it.)
And if the pcb design side is not the main part of your work (If it is expected that you use the hardware you make for more than a basic software demo) then you might want to check if you can find a finished board that holds a processor with enough power. (designing a pcb for a complex thing like a processor of your listed power can easily take a few months of full time work with multiple iterations and therefore additional manufacturing waiting times. And then you need to somehow solder it as well which will add yet another skill to learn.)
Depending on your application either check a fpga supplier (XILINX has nice evaluation boards) or even things like nvidia jetson tk2 (or whatever the current version of it is) if you need a powerful graphics processor. If you already have hardware in mind check if the manufacturer does not offer an evaluation board or at least reference design for it or for something very similar.