Nowadays we have Intel FPGA (prev. Altera), Xilinx, Lattice and other vendors. One of them are better for IoT, another one has better tools and education boards, so which one do you plan to use?
It depend on the allocated budget for the board. You can choose a FPGA board with soft-core for more flexibility, less power consumption and cheaper price or hard-core one as faster but more expensive. The board itself may contain some useful I/O facility such as PS/2, VGA, switches and LCDs... I recommend this board for learning purposes:
I personally believe that Xilinx University Program (XUP) offers the academics certain practical and economical options by which they can improve their research and teaching endeavors. For instance, it can help teachers to adopt a global perspective to identify best- current practices used in educational contexts and by removing unnecessary obstacles, gives the academics more time to focus on more important issues.
I'm partly agree with you, because very often it's all about money. And mostly Altera's boards (I mean terasic of course) are cheaper than same with Xilinx FPGA (by digilent). But, Xilinx boards usualy have so many comfortable pins all around board and also some daughter boards with sensors/connectors/etc so it becomes like lego and very interesting for customization.
But still, I prefer Altera, and I think, that de0-nano board is better than de0-cv, despite that it has lower FPGA family (CIV vs CV) and less peripherals (our students have electronics classes where they can assemble they own peripherals)
Despite that XUP is very good, their instruments (ISE, Vivado) are not very newby-friendly. I suppose, that very good start is Lattice, because their software easy, then student can learn Altera's software - they are very common, but Altera's is nuch more powerfull, and after all that students can be introduced in Xilinx - to see how different everything can be different, especially if it is about SoC-based systems.
I use Xilinx in my course. The main reason is that i am more expert with xiling than altera (now intel), the second reason is that the university program of xilinx provide more boards than altera.
If Altera can provide a simillar offer as Xilinx University Program (XUP) , it would be more useful. In my view, university cirricullum should focus on just one (Xilinx or Altera) and then give a touch of the other one. As in practical field, both types are available in almost equal proportion.