I think its quite in reverse, for Intel to turn Zen into a car crash is going to require over the next 12 months for Intel to slash their prices in ALL market segments especially servers ( think about what Zens current clocks mean for the 32core parts etc).
As a result of that, they would loose far more money then they would if they did nothing, GF/Samsung have only so much capacity. That would smash their stock price and dividend and probably result in the board getting to boot. Intel have 100k employees, 10 of fabs, no real product progression since Haswell and apparently nothing new until after cannonlake. Generally the market is pretty flat, intel are nowhere on mobile or IoT and it looks like NV is running away with AI (until google etc start with the dedicated hardware and then we will see). That leaves Intel with only its already dominate markets of laptop, desktop and Server, only one of those has any real growth. There is no war chest to pay for the war you envisage. What happens if intel slash prices to go to war and AMD still take market share? what does that do to FAB utilization and cost per chip etc on top of the revenue loss?
Then you need to consider that Zen is a small die smaller then P10 (<200mm sq), that AMD single die will address from 4 core desktop to 32 core server. The Second chip (APU) will cover laptop, and upto 4 core desktop. That AMD is very lean (probably under weight @ 10K employee's.) and that Sony and Microsoft have probably paid for a lot of AMD development over the 4 completely different SOC's they have produced for them ( PS4, PS4 pro, X1, Scorpio ). Then end result is a lot of development costs have already been paid (fabric's, interconnects, south/north bridges etc, maybe even some of Zen's core itself). The end result of that is AMD if they need to (intel price war) can price very cheap and still make more GP then they have in the last ~7 years while continuing to use semi custom to help fund development.
Personally i think Intel will do almost nothing until they have to, only if /when its obvious that large market share is being taken. The reason i think that is simple, it gives them the most time before shareholders start to worry which gives them the most time to come up with a response, because at the moment a response of a 6 core in the enthusiasts market and avx-512 in server doesn't look very good. Intel would want to hope AMD dont do a BD to PD style 12 month turn around with 5-10% more IPC and 400 more mhz on base clocks (across the board, consumer , server etc).
You only need to look at the people on this forum who loved to take a dump on AMD and their now almost dead silence (go look at what it looked like this time pre bulldozer) to know that AMD have at least a bright short to mid term future.
Facebook
Twitter