There's several factors making it hard to enter the CPU market besides technical challenges. One is that there's so many offerings already, covering various nic

There's several factors making it hard to enter the CPU market besides technical challenges. One is that there's so many offerings already, covering various niches (embedded, DSP, desktop, mobile, server, GPUs), all with existing toolchains. Another is that accessing a fabrication plant is extremely expensive/difficult/lengthy. Then there's all the IP and lawyer issues. And finally, I'd say that CPUs are mostly a "solved problem". Can we make better CPUs? Sure. But can you get enough millions invested to develop a whole new family of CPUs and hope to compete with what's out there? That's going to be very very hard.