Double that and then your in the right ballpark.
My question is will they have learned anything or will it be a return too stagnated production of 5% per year when they do.
Cause between sandy bridge and 9th gen there was no leaps at all in my view and while i honestly loved the 2500k i had no option in my mind until ryzen 1700 and if intel doesn't see the flaw in that they are doomed to repeat it a massive issue to them.
That's not what happened nor is it what's happening!
If you remember when Ryzen was at it's infancy everyone was talking about TSMC's 7nm node and how it would be a game changer and all, around this same time Intel was boasting how their 10nm process was better then what TSMC's 7nm would be, they had higher density and better clocks and all that.
But you might also remember that Intel's 10nm node was already delayed by years at this point! And it kept being delayed untill TSMC's surpassed it, and then it didn't matter anymore and Intel just gave up on it, changed their naming scheme and moved on to their next node.
You see the problem wasn't that Intel didn't want to make better CPUs, the problem was that way back when, they decided they would try to make better nodes on existing machines! Because they didn't want to spend on new ASML machines, and that's literally all there is to it! They took a risk believing they could make better nodes without buying new, more advanced lithography machines, and they actually eventually did it. It just took so long and cost so much and had such terrible yields that it didn't matter in the end.
But now Intel has learned it's lesson and they secured all the next-generatin High-NA EUV machines from ASML for the next few years (They literally acquired everything ASML could possibly produce and with this move Intel has taken TSMC out of the game for at least a few years)
Now TSMC can still bring new nodes with the machines they curretly have but once intel has High-NA EUV up and running they will have an advantage.
You see, what actually determines hardware performance isn't the architecture, it's the process node, if you want to make a big difference with architecture alone, you will need many years improving little by little untill you reach a meaningful improvement.
The only thing AMD brought with Ryzen that makes you feel like there's plenty room for improvement is a really large stack of products you can go from a 4 core Ryzen 3, all the way up to a 16 core Ryzen 9, and you also have x3d around. But if you actually look at the performane improvement from generation to generation it isn't actually much better than waht Intel had back when they were dominating and it's mostly what's expected from the node process jump, with little in the form of architecture that's becasue they where just moving slowly throught the not very meaningful recent process nodes from TSMC. Just the same as Intel's back then they were stuck with their existing machines, each node were just slightly better than the previous one, to the point they stopped even giving them new numbers and just started placing +++++ in front of it.
Had intel bought EUV machines and started planning EUV nodes back around their 5th gen, they would be in a way better position now, but that's what happened, it isn't that they didn't want to move on, they just couldnt, becasue management told them, they wouldn't get shiny new toys to play with and make better CPUs.
Also, I'm not so sure AMD's Strix Halo will do much, they are already delaying it, they might just make a paper launch of it, AMD really seems to be leaving the consumer market behind, especially on mobile devices, they really do'nt seem to care, their current lineup is leterally worse than the previous one, and companies are complaining they don't get much volume of chips from AMD, and that's ignoring the fact that Stix Halo won't necessarily beat Intel's Core Ultra 200s, they are beasts, I've been talking with friends for years that eventualyl we would get RAM chips in the CPUs for better integrated graphics performance, and altough Apple did it first, Intel has taken the lead and brought it to the windows machines before AMD, while bringing their new Xe architecture as well, so I predict Intel to win this one, even if they lose the power batte they will still win becasue you'll actaulyl be able to buy the intel laptops, while AMD's Strix Halo machines will be too expensive and forever out of stock.
Oh yeah and evidencing how AMD isn't actually trying too hard to improve Ryzen, Intel has a new interconnect ready while AMD is still struggling with Infinity Fabric that was clearly limited since the first iteration, back then I tought they would soon introduce a new solution but no, they are just patching it up bit by bit, intead of developing a better solution, I mean, they were first to market with some kind with 3D chip implementation, but they just seem to have stopped there, at least I haven't heard of any new AMD scientific paper on the matter while Intel is full steam ahead with 3D interconects.