TSMC Completes 5 nm Design Infrastructure, Paving the Way for Silicon Advancement

only idiots said that
we pretty much concluded that intel was lying the day they refreshed the skylakes after the first ryzen release
and pretty much confirmed that their 10nm is shit with the cascade lake fiasco

aka
their 10nm was just a die shrink and nothing else

BASED LISA. Can't wait for my 5nm Ryzen 3 APU laptop with 12 threads 3.8GHz and 5W power consumption that makes the battery last 4 days.

Attached: 1544416271477.jpg (757x627, 87K)

NOOOOO

Attached: 1527629778452.jpg (679x758, 54K)

just delid and delap it

>laptop
*thinkpad

Intel will get their shit together

Only delusional AMD fanboys think otherwise

In 10 years they'll reminisce how Ryzen was kang and how they were forced to use Intel's next Nehalem

>Ryzen 5000 series
>5nm
>DDR5
>PCIe 5
>5GHz all-core
AMD better deliver that 5-5-5-5-5. Plus Arcturus or whatever they end up calling it.

Zen2 isn't even out yet bruv

Obviously this is an exaggeration but I'm genuinely excited to see whats in store for zen2 APUs, considering the leap in performance from pre zen APUs to now

but a ryzen 5 3rd gen beat an i9 9900k bruv

A friendly reminder to all these retards jumping into conclusions.

Intel's 14nm is comparable in area to TSMC's 7nm while Intel's 12nm is to TSMC's 5nm.
Intel's 10nm is bleeding edge tech and no wonder they have so many trouble with it.

Attached: 1514399978661.jpg (1920x1266, 515K)

Attached: 1535454693683.png (600x600, 9K)

Makes no difference whether it's +/- 5% of 9900k, it's still going to outsell Intel.

NO goy ples buy intel

>Intel was sandbagging 10nm just to BTFO AMD when Zen originally dropped in 2017
lol

Who cares what nm it is. What matters is performance/price

So Intel is finished.

I really dont like jews so im rooting for you, shame there are 8 intel computers in this household. I guess ill buy an amd next time

Will intel next products include the vulnerability features that have existed for 20 years?

No goy you're perfectly safe.............

but real men own fabs sir
please do the needful

Different CEO

Working as intended, goy

You can't be serious

He's waiting for Intel 5nm

it's a feature, goy.

This time however Intel does not have neither node advantage or a new architecture that can either reuse or improve. How are they supposed to make a comeback if they can't make 10nm work after 5 years?

yeah they'll suddenly gut their billion-dollar fabs and outsource everything to TSMC. dumb fuck

Sure, let's comprare a working 7nm and a next-to-work 5nm with a "may exist" 10nn

just wait goy

Not that user but myself I already decided to wait for Ryzen 4000 since I want a 7nm APU and they're not coming this year

Or perhaps even Ryzen 5000 and AM4 successor because an APU could really use some DDR5

He was comparing working 7nm with a working 14nm though, 10nm was being compared to 5nm and they're both a future thing, and you have to be truly delusional to imply 5nm status is ahead of 10nm

none of these advances hold a candle to 3d xpoint which can be fab'd straight onto the die
communication, not processing, is the bottleneck in modern systems

>you have to be truly delusional to imply 5nm status is ahead of 10nm
No one has ever been such an Intcel to say something like that. 7nm Is comparable to the original Intel 10nm, but given that Intel is reducing his node to finally use it, it means that it is probably behind TSMC 7nm. Now imagine that EUV 5nm will bring, it is in risk production NOW and probably ready for next year. Intel is not even producing anything outside a few useless i3. Guess which process is really behind schedule?

so what happens when they can't shrink transistors anymore? they can't possibly stack silicon dies forever, right?

moar coars!1!! and quantum co-processors

N-Not fair. B-B-but Moore law ended in 2009 when we introduced Sandybridge. Why is AMD trying to lie? REEEEEEEEEEEEEEE

maybe they can, if they can fit cooling in tall stacks of silicon. or maybe just add cpu sockets. so an average person can have dual socket for his desktop

based chinks

Except Intel hasn't been able to right course for over 5 years now. Remember a Haswell refresh wasn't planned let alone 4+ architectures on 14 nm. 10 nm was suppose to arrive in 2016.

Attached: 1534422888333.png (1260x892, 55K)

10nm might be dead in the water but they're working on getting back on track with 7nm EUV. Whether 7nm is enough to make a comeback remains to be seen but it's not like they have no prospects at all. Also maybe that retarded 3D stacked die sandwich cpu amounts to something eventually. Never know.

that's what happens when you hire h1bs

Yeah I can't wait to upgrade my pc to an arm processor

ARM would be a downgrade

>3D stacking huge, power-hungry dies with sky-high clocks
It will be the end of us all

why do you hates divertisty sir

it would save a lot of power

>tfw ryzen + nvidia
I for one welcome our asian overlords.

2077 is finally the year intel 5nm superpower will destroy amd

Not really
Even in phones the only reason they are so low power compared to laptops/desktops is because of all the fixed function hardware and the fact they are just weaker

I do agree that we may eventually see a 7nm process from Intel eventually, But given the amounts of delays, money spent and lyies about how broken it was, it is hard to believe in Intel once more. This even makes me wonder how Intel after 5 years is still without a demand for lying.

delit this now sir

Attached: intel shlomo.jpg (691x771, 112K)

Has anyone got a chart of scaling factors compared to each companies process node? What makes AMD's 12nm nearly equivalent to 14nm on Intel for instance?

based incel poster

>so what happens when they can't shrink transistors anymore?

Other materials. Maybe graphene, but that one is still a meme - I want my 1000GHz processors already.

Yeah it needs some material that has less resistance.

itproportal.com/2010/02/08/ibm-debuts-100ghz-graphene-processor/

What happens after 5nm?

You get quantum effects where electrons flip and give incorrect readings.

what are they going to do to improve the processors?

More cores

Guy's fucking clueless. Intel's 3D stacking is for ultra low power parts. It's not competing with Zen 2 on any level. If anything it's competition for Snapdragon laptops. Neither stands a chance against ARM Macbooks.

Different metals and coatings to prevent crosstalk between the lanes or something. It may not be on silicon at all though. Perhaps biological is one way forward using DNA strands.

3nm GAAs from both TSMC and Samsung, with the rest of the industry following 2-3~ years behind as every upcoming Chinese fab inevitably steals the IP. Maybe Global Foundries will continue with their hard push for UTBB/ET-SOI processes past 12 FDX, if thats the case then there will be absurdly cheap and high performance processes available at prices we haven't seen for years.

There is still quite a lot of life left in silicon substrates and conventional transistor topology.

DNA computing is a meme invented to secure research funding from suckers. Its going nowhere.

>more cores
Brainlet here. How will they fit more cores into the same chip if they can't make everything smaller anymore?

Attached: 1554391187963.gif (500x281, 426K)

Make the chip bigger

Moore cores past a certain point gives diminishing returns in most workloads. There will be a day when we are measuring our e-peens by the number of 3D layers that are stacked on our processors.

We need an optical computer. Light speed is flippin fast.

Microchip size doesn't really matter anymore, they're already as small as can be.

What really matters these days is battery size, power consumption

>buying a 3D stacked meme chip instead of a God tier vertical GAA optical interconnect chip
Wasting your money, fren

Theres been tons of research into optoelectronics, optical interconnects to replace copper for signaling, even optical logic gates.
There gets to be a scaling issue with photo receptors because of the nature of them. You need a fairly long tunnel to ensure the photon has high enough probability of hitting something in the receptor to produce a charge. It'd work for signaling, but not replacing the xtors in a super dense chip.

>not understanding why smaller processes contribute to lower power consumption
Why are you fucking here?

First person to invent a nanometer scale optical resistor wins the internets.

Thats what AMD did and look at them now

i-intelbros?

Its over. Time to sell all the stock and retire to Israel. What a shoah. Anti-semitic Micro Devices is killin us.

>"look at them now"
>in steady decline since Barcelona, lost 90% of their value
>bound by a shitty contract with a good-for-nothing foundry that would be even more useless if they didn't license everything from Samsung, still managed to fuck up their magical 5,000 niggahurtz 7nm
>endless GCN respins that cannot compete with 75W NVIDIA GPUs
>reached rock bottom in 2016
Nobody expected AMD to survive, but Intel needs a fucking miracle for it to work

>stacking

he is actually making a case as to why intel for the past 15 years hasnt delivered a 10nm as their own slides promised..