Will be interesting to see the Intel response.
AMD is doing really well.
AMD is doing really well.
That, at first, will likely seem like the "rational" way to balance this "new competition" from AMD with keeping investors happy, but they're forgetting one thing -- in that equation the "brand inertia" they love to take advantage of is going to erode with each such "minimal response", until there is none or very little left.
What I'm trying to say is that consumers will put up with a company releasing sub-par/low-value products because they "trust the brand" only for so long, before they give in and start embracing the competition's brands -- as they should.
I mean, unless you, as a consumer, suffer from the Stockholm syndrome, you shouldn't be rewarding Intel for being forced to lower prices on some of its products or add more cores, just like you shouldn't have rewarded Comcast for offering fiber in places where Google Fiber arrived. You should be rewarding the competitor that caused that to happen -- that is if you'd still like that competition to continue in the future.
In Google Fiber's case, that competition disappeared because people were unwilling to reward it and stuck with Comcast/AT&T. And now they'll suffer from it for another decade or so, until another major disruption/competition appears (SpaceX satellites maybe?).
Intel has created a situation for themselves where they seem to have forgotten how to care about the purpose of innovation.
Also, I can go on CDW right now and get an Epyc server, no problem.
Availability isn't an issue for AMD.
How do you know AWS/Azure etc. aren't using them just for price haggling with Intel, as it was done with AMD in the past all the time? The fact you can get EPYC servers doesn't mean they are wide-spread anyway.
That is just unrealistic expectations in the server space. These aren't consumer products where adoption is fast. Business don't upgrade their infrastructure as quickly as consumers, and when they do decided to upgrade it is months of planning. No company is going to jump ship to AMD when their current servers aren't fully depreciated by their accounting standards, and they still have several years left on their support contracts.
The EPYC sales will come, but not overnight.
EDIT: And I am not sure why are are so disappointed here. Epyc sales and adoption has been in-line with what AMD has given as guidance. Why would you expect adoption to wildly exceed AMD's own guidance? I think AMD had aggressive but realistic guidance and so far Epyc has been great success and will only continue to chip away at Intel. Also not sure what revenue miss you are referring to. Overall AMD beat their expected earnings per share by 1 cent last quarter. If there was a slight miss on the Epyc sales then they made up for it somewhere else, but it must not have been a very big miss otherwise they would have missed the EPS target.
2% market share for EPYC. Do you think I bought Threadripper to spread FUD about AMD?
On tech side, AMD is going for continuous increase in density. A lot depends on cooperation of memory makers. Cheap HBM2 or HBM3 on package with CPU will probably the only thing that will endanger Intel's position.