Yeah, basically the first of Intel or AMD to remove all unremovable closed source magic that has root access to your entire system, could become the dominant player for years.
(disclaimer: I work for IBM)
I think I saw some OpenPOWER machines on IBM's website, but all the distributors have stopped selling them. Seems impossible to get hold of an OpenPOWER motherboard :( But yeah as soon as I can buy one I'm getting one to try it out.
Mind you does Linux even compile on it yet?
As of POWER8, we now support running Linux bare-metal, and we sell machines with Linux bare-metal, Linux as a guest under PowerVM (our proprietary hypervisor) and Linux as a guest under PowerKVM (which as you can probably guess, is just a distribution to ship an open source KVM/libvirt hypervisor). https://www-03.ibm.com/systems/power/hardware/
No BSD support as yet ;)
Edit: Ahh okay I see. The higher end ones need you to call IBM, but I think you can just buy the 6k one.
The processor could benefit from crowdsourcing development to some extent.
Also, in countries that aren't America (and for many entities in America) there could be huge value assigned to knowing with a higher confidence that your vendor isn't handing over a backdoor to three-letter agencies
Finally, FLOSS enthusiasts will go crazy for it, which might not be a huge boost to AMD/Intel but is at least something.
It may be an indirect way to get more talented and idealistic individuals going to AMD versus Intel.
I don't know much about the subject, so these are all my guesses as to the biggest advantages. Who knows if they outweigh the perceived disadvantages from their perspective.
AMD or Intel processors also definitely aren't good candidates for crowdsourcing. Their R&D probably spends more on coffee than the largest crowdsourced projects raised.
That includes the option for developers to improve the limitations of a platform they're building on top of instead of working around them. My thinking was that the people who would want to do that when it comes to an open processor would be kernel developers.
However your point is good about the R&D, and I bet Intel/AMD spend a LOT of money/resource on those backdoors. But the open source world has done a lot in the past. Linux is a massive open source project and arguably very successful an innovative.
But I do get your point, at this stage it's still a smaller percentage of people who care vs those who don't. But if they capture this market early, couldn't they secure their business for the next X years?