I think the most important question here isn’t whether Arm will win servers. The assumption would be there, that the cloud market will continue to grow at its

I think the most important question here isn’t whether Arm will win servers. The assumption would be there, that the cloud market will continue to grow at its current rate and continue to be the valuable market it has been. What scares intel most is that a new market will emerge for middle and high power compute spread more broadly, and this market will grow explosively faster than data center ever did. The question is whether these new cores will bring about a new way of structuring the data center, where it is decentralized and spread out more instead of stuck in one building. A good example is the infrastructure that must go into place to support 5G and self driving cars. Every lamppost will have a server on it. Arm winning servers becomes irrelevant in that world. If fog / Edge / infrastructure combo blow up the way mobile did, Intel’s little data center slice won’t matter. Developers will only adapt when there is a big new market to write software for. Data center by itself not sufficient to motivate the change needed.