You know Intel's 10nm cpus will stomp all over Zen 2 cpus right?

You know Intel's 10nm cpus will stomp all over Zen 2 cpus right?

Attached: intelinnovation.jpg (1510x849, 104K)

my understanding is that intels 10nm will be slower than current 14nm because of how fine tuned 14nm is. also, its supposed to release in 2020 or later. so intels 10nm will compete against zen4000.

I licked a negro once

Well yeah, sure but Zen 2 will be a 5 year old architecture at that point.

>zen 2
they will be competing with ryzen 4000 next year.

Even Intel themselves admited that 10nm will be worse than 14+++++++

Attached: 10 is less than 14++++++++++++++++++++++++++.png (523x509, 62K)

Attached: 1562549379804.png (680x485, 516K)

ummm no sweaty sorry

Attached: image.jpg (605x644, 81K)

You're right but the poverty-tier amd fanboys won't admit it. These tards actually believe that Intel will never produce a 10 nm cpu.

In b4 "Intel's own roadmaps don't count, AMD BTFO!".

>d-don't worry, Intel will surely release 7nm in 2020
>I-I mean they were on track for 2025
>Intel w-will be k-king with 7nm in 2030!

intel 10nm superpower 2022

It's funny that people keep talking shit about AMD's current performance on 7nm but it's like every other company is incapable of releasing stuff on smaller nodes.

Attached: orig[1].jpg (2048x1490, 292K)

10nm 2022

people are SO MAD LOLOLOL

There won't be a 10nm CPU

They'll leapfrog to 7nm

The next *Lakes are all 14nm

Yes. Node shrinks below about 20nm are particularly not very good at actually increasing performance. This has been an open secret that people are glossing over. To get nodes that small you need to layer your dies so the transistor itself is built in 3d space rather than 2d space, so while the area goes down, it's building upwards, increasing the volume. We should be referring to dieshrinks by the volume of the transistor, not the area. Imagine referring to the space inside a skyscraper by talking about its ground floor area.

in clockspeeds, yes

By the time Intel's 10nm is out, AMD should have something competitive ready or already released.

5nm Zen 2+ will be out by this time.

Attached: amd_running.jpg (680x485, 279K)

Clockspeeds is literally all Intlel has at this point. And memory latency. When they move to a chiplet design to be able to compete on cores, then latency will take a hit too.

The JUST from both sides in this thread.

Attached: a13.png (531x376, 283K)

>Intel's 10nm
>actually releasing 10mm
Wake me when its happening
Also if they are going to try to offset their 10mm clusterfuck costs then the new chips will be expensive as fuck.

2025 is just arond the corner

*benchmark does not reflect vulnerability patches

Maybe AMD should have waited too.

...

They'll delay 10nm to 2020 this year.
Then next year they'll delay it to 2021.
Then in 2021 they'll delay it to 2022.

Maybe in like 2030 when it finally comes out it'll beat the 7nm Rome chips.... after everyone has already upgraded to AMD's chips 5 generations ahead.

Can anyone confirm this?
Sounds interesting desu

Attached: 1517719598776.gif (728x720, 550K)

Intel already released 10nm parts. They're utter garbage which is why they went back to 14++++++plusplus
Intel 7nm will be their next competitive node and that will be competing with Samsung 5nm. Unless Samsung completely screws the pooch like Intels 10.

This, it's basically like boasting that Zen stomped Haswell, obsolete architecture at that point.

Attached: 1488923970547.png (675x482, 486K)

Perhaps aside from good efficiency at moderate performance 7nm isn't really that good for high performance products...?

no it wont
intel has emib

Samsung will make Intel GPUs, they just don't want Nvidia screwing them over.

If the gook made Intel GPUs are a success they'll move their desktop CPUs to Samsung as well until they get their 7nm and beyond right. They will never give someone else their Xeons and mobile chips.

>lmao just port your arch to our process bro
no