If you develop a new CPU architecture from a white paper, without limiting yourself to backward compatibility with x86, how much faster will it perform?

Question

Processors

If you develop a new CPU architecture from a white paper, without limiting yourself to backward compatibility with the x86 architecture and programs written for x86, how much faster can this new architecture become?
In other words, how much backwards compatibility with x86 and the accumulated baggage of x86 software limits the performance of computers?
Could there be a need in the near future to create new processor architectures without backward compatibility with the existing ones?

Answer 1

From 2 to 10 times (this is my value judgment)))).
It can't because RISC-V is available.

Answer 2

If you develop a new CPU architecture from a white paper, without limiting yourself to backward compatibility with the x86 architecture and programs written for x86, how much faster can this new architecture become?

In what tasks?
Backwards compatibility = universality. Lack of versatility = highly specialized solution. Designed from the ground up, a processor for broad tasks will immediately acquire backward compatibility. The result will be the same x86, only side view.
No, it seems to me. If someone is not satisfied with x86, then he has too highly specialized task. Therefore, highly specialized processors begin to cut.

Answer 3

such an abstract question. It's like let's develop a new engine for cars.
If I do this, it will probably turn out badly, much worse. If the engineering department of some Ford, then
hell knows, they probably get something better.