Yet. Their GPUs were not great for years, but have managed to catch up (and even overperform for their price point), so other workloads are yet to come.

Unfortunately with the way ROCm is developed, and how its only intended for specific GPUs, I doubt it.

Still waiting for rocm support on the 5700xt which they kept promising was ready any day now.

Did they ever promise that?

At the moment maybe, but owning the CPU may allow for better integration.

Imagine if AMD launch a new bus implementation from CPU to GPU. That's not something Nvidia can do by themselves. Maybe Nvidia buys Intel and does it though!

You mean a proprietary connection standard? Like G-Sync but for accelerator cards? ;)

https://www.tomshardware.com/news/amd-infinity-fabric-3

AMD already does that, and much like G-Sync there is also an open standard (CXL) that everyone else is converging around.

https://en.wikipedia.org/wiki/Compute_Express_Link