It is really funny when you find out that Intel uses Xilinx FPGAs for prototyping as they cannot get what they acquired (Altera) working in house to make things work.
I don't work for Intel but I do work for a semiconductor company. While Xilinx FPGAs aren't directly used for prototyping, there are a large number of third party boxes purchased to accelerate hardware simulations and they're chocked full of FPGAs.
That's the more likely explanation. Altera vs Xilinx isn't just the hardware, it's an entirely different toolchain. It would be insane of Intel to demand third parties to move all their technology over to Altera's.
I imagine they are using a 3rd party prototyping solution like HAPS from synosys, which uses Xilinx FPGAs inside - for good reason, for quite some time Xilinx have had some very large devices built specifically for this market. It must sting a little bit though....
Xilinx are very flexible and the tooling makes them even more effective.
If AMD wants to get serious in the datacenter/AI/ML space they need a xilinx-like approach to developing tooling. Cuda, nvenc, cudnn etc craps all over amd's offerings in the same space where they are even available.
AMD is prepping to take over datacenter terf and this puts them in a good place to bring a bigger offering.