An interesting observation is that both Turing and Volta can't really get any bigger anymore, they're already at the limits of what's possible in terms of lithography. Turing does not have the FP64 performance for the datacenter market though, and there's no room to add more transistors, so one hint here seems to be that NVIDIA will maintain two different architectures going forward.
In fact, I cannot see either FP64 or raytracing going anywhere any time soon. As such, it’s my assumption that NVIDIA will maintain two different architectures of GPUs going forward. The only way that I can see this changing is if they figure out a multi-die solution, because neither design can get any bigger. And even then, what workload would it even perform? (Moment of silence for 10km x 10km video game maps.)The codename "Ampere" circulated quite a lot earlier this year, perhaps that's the compute-centric successor to Volta?