theinquirer.net
Charlie was right all along.
People were unironically shilling that icelake would be 2h 2018 and would be in time to compete with Zen2. Poor fuckers.
Intel 10nm is borked
Parker Myers
Chase Sullivan
10nm never.
Andrew Bailey
they simply cant just do a die shrink to compete against ryzen
because simply they will have to push the 10nm chips to get the 14nm chips speeds and they will have to create a housefire line
Jordan Morris
They won't have new architecture ready until 2022. Bingbus forever
Anthony James
so even the shitty ULV garbage won't be released on time for zen 2? where is the famous fab advantage intel had over everyone?
Blake Lee
between Ryzen conquering Windows notebooks and Apple going with their in-house stuff Intel's business is going to plummet by 2020
Matthew Stewart
>People were unironically shilling that icelake would be 2h 2018 and would be in time to compete with Zen2.
Yes, the thread where our resident shitposting shill had a full nuclear turboautism meltdown. rbt.asia
Austin Gutierrez
That's what you get when you hire based on """diversity""""" instead of merit.
When will cuck companies learn?
youtube.com
youtube.com
youtube.com
youtube.com
James Kelly
The other companies collaborated, it seems the collaboration paid off because they've either caught up or are ahead of Intel now.
William Peterson
meanwhile at IBM
Jaxson Nguyen
>88
annuda shoah confirmed
Christian Russell
is this 7nm?
David Anderson
Damn, they're fucked.
Adam Collins
no, 14nm, some node only IBM uses that is better for large dies
Thomas Morris
It's called 14HP, will be used on POWER9 too
Camden Brown
Their fab advantage was literally the only thing they had going for them, and now they've pissed that away too.
B R A V O B R I A N
Leo Lopez
what the fuck guys
en.wikichip.org
Jack Cox
>16,192 cores
Sounds like a GPU to me.
Adrian Howard
>>Charlie was right all along.
Of course he was, he was saying so since 2016
Dylan Reyes
POWER10 for 7nm, latest POWER chips are always a generation behind cutting edge.
Though nobody really knows how the 14nm SOI performs
Aiden Richardson
FUCKING KEK
this is what you get by forcing diversity into your company. yep, keep replacing smart senior engineers with retarded """"gender fluids"""" with 850 ancestral lines in the name of diversity. you are doing good. lmao
Christopher Torres
Charlie is usually right.
Oh BOY.
BOY.
NO.
GloFo just announced 12LP to shit on 14++.
Juan Wilson
Boy this fuckup will make even crapple switch to AMD.
Parker Ortiz
14nm++++++++++++++ is b-better than 10nm!!!
Ethan Cook
Who here 10nm?
Dominic Bennett
Who gives a shit.
Intel cores are still faster. They'll just start gluing cores together like AMD and maybe reduce prices a bit.
Brayden Murphy
>They'll just start gluing cores together like AMD and maybe reduce prices a bit.
Developing interconnect and protocol for that takes a little bit longer than that.
Since it took Intel 5 years to copy HyperTransport i guess they'll have their own IF in 5 years.
Julian Moore
>JUST ADD MOAR COARZ
That's not how that works Brian, if your 6 core is a deciding factor on whether or not you can get home insurance...you can't just slap more cores in and expect it to work.
Ryder Fisher
>640W
>740mm2
Fucking hell...
Gavin Nguyen
you forgot the 50TFLOPS FP32
Daniel James
Probably a specialized GPU
Jose Myers
A manycore processor.
Remember Polaris (as in Intel's Teraflops Research Chip)?
Jeremiah Edwards
And the 512 PCIe 3.0 lanes.
Christopher Nelson
Dayum son these chinks are crazy
>120 directly connected NVME drives