Wait for 3nm, don't give them your money

Attached: 1530560464524.png (713x608, 551K)

I'm sorry but I only use GenuineIntel 14nm+++ CPUs.

3nm chips will have massive long-term failure problems.
This current gen is likely to be the last generation of chips (plus 7nm EUV) that don't fail in the first 5 years of their lives.

>wait 3+ years to upgrade your adding hardware

Sorry I'm not poor.

Just wait™

this

>3nm factory
That's a pretty small factory. Where will they find workers who can fit in it.

3nm GAA can't get here soon enough.

Sure, just like 14nm FinFET when you room temp IQ groids made the same totally ass pulled predictions over half a decade ago.

>2012: i5-2500k, 8GB DDR3, Sata SSD + mechanical HDD
>2018: Ryzen 7 2700x, 16GB DDR4, NVMe SSD + Sata SSD

Best upgrade plan desu. I'll upgrade once 16-core 3mm, DDR5 systems hit the market.

How do you expect them to work if they can't even fit inside the building?

Attached: 9349981285981.png (1280x720, 414K)

That's why it's in Taiwan.

AMD BTFO
still stuck on ivy bridge ipc with MOAR CORES LMAO

Ah yeah because 3nm is totally the same as 14nm, or even 7nm.
Please come back in another 8 years kid.

>this transistor with less leakage than any transistor ever is totally going to magically break down over time
>muh quatnumbs
Keep your retarded meme shit on reddit.

>upgrade your adding hardware

>this immensely thin wire will never break down!
>what do you mean, electromigration?
it's time for you to go to bed user. the school bus will be there in the morning.

What's more funny, if you think about it, is that TSMC's 3nm fab will be up and running before Intel's can get 10nm right.

>repeating meme bullshit you know nothing about
Keep on trying, reddit. The physical structures on die aren't going to break down until the atoms they're comprised of reach their half life. Meme tunneling is an overblown issue only whined about by people who have absolutely nothing to do with the semiconductor industry.

>Not waiting a few measly weeks for Zen 2

If you think I'm talking about electron tunneling, you're dumber than I thought you were.

I'm still in 45nm and I dont give a flying fuck about (((upgrading)))

Attached: fullonbotnet.jpg (500x342, 82K)

>performance doesn't matter!

Attached: intel given up.jpg (512x599, 131K)

AMD Wait™?

Do you mean GenuineIntel 10+++++4nm?

PRESS S TO SPIT ON INTEL'S GRAVE

Attached: 1536694146733.png (1200x800, 164K)

but user the silicon limit is physics, no matter what industry you're in science is still science

Then explain 5nm and 3nm

wait for intel 10nm in 2077 it will btfo lolol

That's a comfy looking shack

The finite limit to area scaling isn't being reached at industry 3nm with GAA transistors. There is a very long way to go before that.

I am a brainlet, does that mean we'll get μm architectures with dozens of cores for consumer market that are not high end someday?

this. what productive application (besides rendering and video processing) requires newest cpus?

The density and feature size of all the metal layers that carry power and IO to/from the package are still enormous, and there is ample room to improve total density and area scaling there. If front end feature size remained totally static the industry could still produce smaller chips by working on back end features and materials.

There is potential left on the table everywhere.

Faster browsing on bloated OSs?

A larger cache makes a processor vastly faster for memory intensive applications.
More threads makes it faster to run multi-threaded applications or just several applications.
Other than that, most improvements are made in specialized functions like in the GPU for those who use that in the same chip.
Other times, it is simply a matter of improving thermals/power efficiency or clock speed.

The final stage of shitel