Well Tiger Lake in particular won't be competing against Zen3 for awhile. Its main competition is Renoir. So I do not understand the semi-defeatist attitude being exuded by Swan. Do they expect Tiger Lake to lose to Renoir?
Getting ahead of the game, I mean is there any proof that 10nm is in any better situation then last year or the year before. If its not, its still going to be small dies, and bad clocks, and maybe still bad yields. Are they going to have to pump up the power on these mobile chips just to get base clocks up? I mean look at Icelake, 28w for a 2.3GHz base clock 4c part. AMD has a 35w 8c part with a 2.9GHz base clock. Looking at 4u chips, AMD has a 2.7GHz base for a similar 4c 15w unit (even though it carries some overhead of being originally a 8c chip. If tiger lake is all IPC improvements and no clock improvements, even at the growth they are touting, its still going to be a rough ride. At some point Intel is going to have to bite the big one and move their other markets over as well. Is a 10nm Xeon going to hold up against a 7nm+ Epyc? A 5nm Epyc. If AMD keeps up with 15%-20% even if Intel gets 25% a gen its going to be a tough fight once Intel leaves the comfort of 14nm.
But honestly. This is just a seed. It was said, public will keep an eye out, and they will choose when the market actually is ready for No Benchmarks. If TL is competitive, specially if it wins enough for them to make a few slides, then the market isn't ready yet.