Oh no no no, is Intel ready for this?
Oh no no no, is Intel ready for this?
Other urls found in this thread:
MAGA
(Make AMD Great Again)
I will never buy an AMD chip
>World near 20% filled with npc
>AMD's market share is about 20%
Intel is not ready, obviously
t npc
>>World near 20% filled with npc
more like 60%
how much time has amd left in order to buttfuck intel the most it possibily could on the server, workstation, destkop and hpc market? (the notebook one is a lost cause sadly, unless they severely undercut intel on perfomance/price while also offering more battery life)
Intel will have mass production 10nm ready in about 18 months? Is it plausible?
But even with a working 10nm, their core architecture is old, and cant scale well. Their ringbus can't stand a chance against infinity fabric, which also gave amd a massive advantage on the yelds problem.
So, how much time for a completely new architecture that can keep up with zen efficiency and superscalability?
Nice try, but it was listed at 20% you soulless faggot that does what is memed.
Rome will likely ship in late Q2 to the general public. Q1 for hyperscalers.
If intel unfucks their 10nm they may be able to ship ice lake SP sometime 2020 at the absolute earliest. So AMD has at least about 6-9 months to run with the lead. Maybe longer if icelake is not as competitive or there are additional manufacturing problems
I would bet icelake roadmap will end up being different than how we see it now. OEMs and ODMs are going to feel the pain in the roadmap churn and will be reluctant to ship earlier than things are stable. So there is that variable too.
10nm isn't some kind of magic that's just happening. Even assuming Intel is finally capable of unfucking their clusterfuck of a process, it's still worse than their initial process planned for 10nm. And there's no guarantee that it's even on par with their current 14+++nm process performance-wise. Hell, even in their slides a few years ago, their first 10nm process CPU is supposed to be slightly worse than their last 14nm offering. And that's the canned process.
And that doesn't even take into account the shitty yields of monolithic chips while AMD can shit out chiplets like no ones watching. Even if Intel can catch up somehow, AMD is still likely to be the way better bang for the buck until something fundamentally changes at Intel.
From the thumbnail I thought it was a Hillary vs Trump poll from 2016.
>Maybe longer if icelake is not as competitive or there are additional manufacturing problems
Didnt disagree with you in my post
Are they ever going to add modern AVX stuff?
give it a day or two for passmark to (((correct))) it
Because buying shit from a monopoly makes you a based (((magapede))) doesn't it
it doesn't even matter. AMD is proprietary now. no matter how much "competition" there is between the two competitors we might as well be in a monopoly.
>proprietary
Go tell me how IBM power is based despite x86 duopoly being IBM's fault, do it newfag
what do exactly do you think I'm trying to say that you think you're disproving?
That you are an absolute cringe contrarian, you can go back to /reddit/ now
It'll be alright son
I'm just saying I wish hardware was less proprietary.
10nm is meant to be ready in 2020, but I doubt they will be able to make huge dies like the ones in Xeons without getting 1.7% yields.
the only reason intel has such a massive market share over amd is because amd focused on desktop shit instead of laptops and windows tablets, aka shit normies buy in droves.
I would be very interested in a market share chart that only shows desktop cpus.
this is obv a fake. dude just took one set and mirrored it
The original paper stated it was ~20% who had the inner monologue, so it was ~80% NPC
which would make sense with OPs graph, as you would buy a lesser quality product only if you had a moral reason too.
this is changing with zen and zen2 however.