Intel Kills off the 10nm Process

Zarathustra[H]

Official Forum Curmudgeon
Joined
Oct 29, 2000
Messages
29,462
Careful there, there's no real standard for what "10nm" or "7nm" means across manufacturers.
Intel's 10nm process should (but might not) have higher component density than its 14nm,
but Samsung/TSMC's 7nm isn't necessarily more dense than Intel's 10nm.

There is an ITRS standard.

It's just that no one follows it anymore. Intel did for a long time, but even they have started relaxing and not meeting it in the last couple of generations.
 

pcgeekesq

[H]ard|Gawd
Joined
Apr 23, 2012
Messages
1,399
INTC closed up 2.29% on the news.
So "the money" isn't taking SA's report as "news," either because they think it's fake or because they have known for a while.
Probably the former.
 

dgz

[H]ardness Supreme
Joined
Feb 15, 2010
Messages
5,692
INTC closed up 2.29% on the news.
So "the money" isn't taking SA's report as "news," either because they think it's fake or because they have known for a while.
Probably the former.
Intel is immune to almost everything it seems. Doesn't matter what they do. A billion set aside for fines? THANK YOU for that mosquito bite, EU and US. Imma invest my time "innovating" new ways to sell 4 cores and have people drool over them. That went for quite a while, didn't it
 

Pieter3dnow

[H]ardness Supreme
Joined
Jul 29, 2009
Messages
6,789
I have been hearing talk of how Intel would ROFLSTOMP AMD when they came with Zen and just go with 10nm and even before that 10nm was a good reason to not even dare to make anything for years.

If you followed multiple stories around the 10nm fab problems and the contracts Intel had were not going as planned then why wonder that 10nm got axed?
https://semiaccurate.com/?s=10nm

https://semiaccurate.com/2018/07/02/intel-custom-foundrys-10nm-meltdown-is-crushing-a-20b-market-cap-tech-giant/

You can claim all the stories are false but in that respect where are the glorious 10nm products I mean the 10nm products that outperform their 14nm++ counterparts ..
https://semiaccurate.com/2018/08/02/intel-guts-10nm-to-get-it-out-the-door/

https://semiaccurate.com/2018/05/29/is-intels-upcoming-10nm-launch-real-or-a-pr-stunt/

The chip in the Ideapad 330 is a Cannon Lake based 10nm CPU from Intel that has been blessed with the marketing name Core i3-8121U. If you look at the main Ark page this part isn’t currently listed but if you search for it specifically you end up on a spec page. There are a lot of interesting bits to be gleaned from this page starting with the name. 8121U connotes a low end i3, the lowest end of the non-crippled all that much range. Think products that go into bottom-feeder special laptops, then step up half of a notch. Volume, not quality.
In the end we have a chip being built on a troubled 10nm process. In six months Intel can make 15% of a day’s production on those production lines. The resultant chips are abjectly broken, they are 2+1 but the +1 doesn’t work which means they are selling a CPU with half the die turned off, an expensive proposition given the cost of the process and the shatteringly low yields. Even with the GPU turned off, the resultant CPU uses twice the power of the 14nm devices to run slower than those with a GPU.
Why do people find it far fetched that 10nm is not something Intel even wants ?
 

Qthulu

[H]Lite
Joined
Apr 19, 2018
Messages
103
I'm not sure 10nm is really ever gonna pay off for intel. Its not gonna be the cashcow 14nm has been for them.
 

Qthulu

[H]Lite
Joined
Apr 19, 2018
Messages
103
If they did kill 10nm off they would lie about it to protect share prices anyway.
 

IdiotInCharge

[H]ardForum Junkie
Joined
Jun 13, 2003
Messages
13,241
I have been hearing talk of how Intel would ROFLSTOMP AMD when they came with Zen and just go with 10nm and even before that 10nm was a good reason to not even dare to make anything for years.
You should get your hearing checked.

You can claim all the stories are false
You just used SemiAccurrate to validate... SemiAccurate. No claims need to be made.

Why do people find it far fetched that 10nm is not something Intel even wants ?
Intel wants to make money. It's what they're good at.
 

Pieter3dnow

[H]ardness Supreme
Joined
Jul 29, 2009
Messages
6,789
You should get your hearing checked.
You just used SemiAccurrate to validate... SemiAccurate. No claims need to be made.
Intel wants to make money. It's what they're good at.
You meant shoot the messenger ignore the message. Works for you right ;)
Why don't you check what Charlie says about the 10nm product that is already in the shops , you can buy it today then do the mathematics.

I never heard anyone with a fab getting money from really bad yields but if you care to explain how exactly give Intel a call you have a rosy future ;)
 

IdiotInCharge

[H]ardForum Junkie
Joined
Jun 13, 2003
Messages
13,241
You meant shoot the messenger ignore the message. Works for you right ;)
You gonna quote tabloids next?

I never heard anyone with a fab getting money from really bad yields but if you care to explain how exactly give Intel a call you have a rosy future ;)
Which proves SA correct? Bad yields now means that they've conclusively abandoned it? And by conclusive we mean SA made it up :ROFLMAO:
 

oldmanbal

2[H]4U
Joined
Aug 27, 2010
Messages
2,161
so can we safely assume that 90% of news specific to a company are BS aimed at day trading?

I'd imagine they're not even aimed at people directly, but targeting algorithms that competing firms use to make their hyper-fast short term trades.

There's a money war going on and it leverages stupid social media sites to spread misinformation so they can adjust stock prices the way they want for a few minutes (the key being they can predict what will happen and when for that short period of time after a new story breaks).

We aren't going to see a stop of this kind of market manipulation until they make rules that state that stocks must be fully acquired and held for a certain amount of reasonable time (1 continuous 24 hour period) before they can be sold again. Short term trading like you see with the ultra rich (with their insane low latency connections to markets and algorithms that trade thousands of stocks every second) is a cancer to the system and needs to be stopped.
Regardless of your point, Charlie has had a raging hardon for the demise of Intel and Nvidia for a long time now. Even the slightest hint at a delay from either lights his Christmas tree.
 

IdiotInCharge

[H]ardForum Junkie
Joined
Jun 13, 2003
Messages
13,241

jamesv

[H]Lite
Joined
Mar 12, 2016
Messages
102
As the dumbest person in this thread I’m curious about something.
If you keep dropping the ball on 10nm, why is Intel acting like 7nm is some sort of relief?
Wouldn’t 7nm add more problems.

And how is it that AMD can have 7nm already.
Did they make Ryzen so different it can leapfrog over Intel?
 

Nobu

2[H]4U
Joined
Jun 7, 2007
Messages
3,669
As the dumbest person in this thread I’m curious about something.
If you keep dropping the ball on 10nm, why is Intel acting like 7nm is some sort of relief?
Wouldn’t 7nm add more problems.

And how is it that AMD can have 7nm already.
Did they make Ryzen so different it can leapfrog over Intel?
Basically, they've been developing these different architectural nodes (12nm, 10nm, 7nm) in parallel, and had a set schedule to complete them, begin testing, ramp up, and volume. Well, 10nm isn't working out too well, and is way behind schedule. Their 7nm design is getting near completion. There is the option to just scrap 10nm (call it a loss) and redirect funds which were going to 10nm development toward 7nm. The smaller node would save them money compared to 10nm, and they may not have even used 10nm for long due to how far behind schedule it was. If 7nm was designed around different fab technology than 10nm (likely), then issues with manufacturing and yields may not be as prevalent.

That is all assuming Intel's 7nm is close to usable, and ignores probably many factors that they would want to consider, but that's the thought that goes behind that idea.
 

Johan Steyn

[H]Lite
Joined
Dec 1, 2016
Messages
85
Charlie responded to Intel's denial and he said that he stands by his report, basically calling Intel out as lying about it.
 

Mr. Bluntman

[H]ardness Supreme
Joined
Jun 25, 2007
Messages
6,408
The article is basically worthless without the analysis, which is held ransom behind a $1K/yr. paywall. When the wall went up I stopped reading. It was different when he was freely available to read on the Inq...
 

jamesv

[H]Lite
Joined
Mar 12, 2016
Messages
102
AMD doesn't- they don't own fabs.
I misunderstood AMDs Presentation obviously.
OTOH I’m enjoying the competition AMD has caused.

I’m running a small recording facility and building stage worthy 1U PCs.
I remember when Tyan Tiger mainboards with Dual Athlon MPs ruled.
Glad to have them back.

I still have to use the fastest single core performing CPUs regardless of cost.
So Intel has been winning that for years now.
Would love to see AMD Zen2 CPUs with big IPC gains.

Great info in these threads.

Thanks
 

akaliel

Limp Gawd
Joined
Feb 7, 2007
Messages
210
The article is basically worthless without the analysis, which is held ransom behind a $1K/yr. paywall. When the wall went up I stopped reading. It was different when he was freely available to read on the Inq...
Charlie's paywall is $1000 a year? lol, omg.
 

drescherjm

[H]ardForum Junkie
Joined
Nov 19, 2008
Messages
14,641
That's less than most cable bills, but still outrageous. Edit: well about equal to a cable bill ($83/mo) before hbo, etc.
I wish my cable bill was only $83 and have no premium channels at all.
 

N4CR

[H]ardness Supreme
Joined
Oct 17, 2011
Messages
4,439
Lol the same Intel said 10nm was all going on track and then it was all of a sudden delayed, then they had 'inventory product, improving yields' etc etc after they launched those 10nm leonovo lemons this year. (On a side note I'd love to buy one as they will probably be collectable one day as the ode to 10nm being a failure). Those Laptops regressed in all performance aspects (less clock, more power, less speed) and had a worse than 50% yield as they couldn't even enable the damn iGPU or get better than a dual core out of a quad die lol.

I'll believe Intel's denial when we actually see these fabled 10nm products.
As above, they did launch a 10nm laptop... barely. If yield was any worse it would've been a single core!


I'd say skipping to 5 or 7nm is a better move at this point as they're about to be left behind at the end of this year. But Intel has it's own' fabs.. so doing that is far, far more difficult than hitting up TSMC for a few libraries and wafer starts.
10nm and how it is managed may make or break Intels image for the next 5+ years.
 

dgz

[H]ardness Supreme
Joined
Feb 15, 2010
Messages
5,692
You said, and I quoted,



I responded with their CEO getting fired:



You responded incomprehensibly whining about stock prices:



And I'm still wondering how a company that is 'immune' loses their CEO.
Ah, my bad. I meant immune in the context of wall street
 

jamesv

[H]Lite
Joined
Mar 12, 2016
Messages
102
Basically, they've been developing these different architectural nodes (12nm, 10nm, 7nm) in parallel, and had a set schedule to complete them, begin testing, ramp up, and volume. Well, 10nm isn't working out too well, and is way behind schedule. Their 7nm design is getting near completion. There is the option to just scrap 10nm (call it a loss) and redirect funds which were going to 10nm development toward 7nm. The smaller node would save them money compared to 10nm, and they may not have even used 10nm for long due to how far behind schedule it was. If 7nm was designed around different fab technology than 10nm (likely), then issues with manufacturing and yields may not be as prevalent.

That is all assuming Intel's 7nm is close to usable, and ignores probably many factors that they would want to consider, but that's the thought that goes behind that idea.
Thanks for explaining.
The 7nm I cited that AMD had was a misunderstanding, seems that was a GPU.

But isn’t that kind of like prepping for your CPU?
Seems logical to develope something like a GPU before going all in?

Exciting stuff though.
Since Intel made record revenue in Q3 I won’t feel bad switching to AMD for my next build.
I’m still getting everything I need done on a Quad i7 as long as it’s above 3.4GHz.

The Zen2 will have a GPU and most likely a Quad at the bottom of their choices.
Last time I Used AMD was when the MPs were rolling right over Intels line up.

I sure love the competition.
But AMDs single core performance needs to be equal or better if they want real competition.

Thanks

Cheerz
 

Nobu

2[H]4U
Joined
Jun 7, 2007
Messages
3,669
Thanks for explaining.
The 7nm I cited that AMD had was a misunderstanding, seems that was a GPU.

But isn’t that kind of like prepping for your CPU?
Seems logical to develope something like a GPU before going all in?

Exciting stuff though.
Since Intel made record revenue in Q3 I won’t feel bad switching to AMD for my next build.
I’m still getting everything I need done on a Quad i7 as long as it’s above 3.4GHz.

The Zen2 will have a GPU and most likely a Quad at the bottom of their choices.
Last time I Used AMD was when the MPs were rolling right over Intels line up.

I sure love the competition.
But AMDs single core performance needs to be equal or better if they want real competition.

Thanks

Cheerz
Np. Note that I was just using those numbers as an example–I don't remember off hand what intel has coming after 10nm, and before 10nm intel had (has) 14nm(+++), not 12nm.
 

jamesv

[H]Lite
Joined
Mar 12, 2016
Messages
102
Np. Note that I was just using those numbers as an example–I don't remember off hand what intel has coming after 10nm, and before 10nm intel had (has) 14nm(+++), not 12nm.
Yes but your explanation of the process enlightened me.

Nice to have a place where I can learn without having to research.
Something that isn’t of major concern being explained succinctly saves me from reading 50 different lengthy explanations.

Cheerz
 

juanrga

Pro-Intel / Anti-AMD Just FYI
Joined
Feb 22, 2017
Messages
2,562
If Intel can't do 10nm, how they going to do anything smaller?

So for now its only TSMC and Samsung that can do 7nm.
As the dumbest person in this thread I’m curious about something.
If you keep dropping the ball on 10nm, why is Intel acting like 7nm is some sort of relief?
Wouldn’t 7nm add more problems.

And how is it that AMD can have 7nm already.
Did they make Ryzen so different it can leapfrog over Intel?
The 10nm node used by Intel on CannonLake chips has a density of 100.76 MTr/mm². The 7nm node used by Apple has a density of about 80 MTr/mm². The 7nm node used by AMD for Zen2 has a density of only 67 MTr/mm².

So the '7nm' node will be used for Zen2 is about 50% bigger than the 10nm node is used in CannonLake chips. That '7nm' node would be 12nm per Intel foundry standards.

Next year Icelake is expected to use a 10nm+ node with relaxed parameters compared to 10nm, to give higher performance. Using the 14/14+/14++ relationships, we can guess that 10nm+ could be about 91.6 MTr/mm².
 
Last edited:

juanrga

Pro-Intel / Anti-AMD Just FYI
Joined
Feb 22, 2017
Messages
2,562
Charlie Demerjian of Semiaccurate has dropped a bombshell article in which he reports that Intel insiders have proclaimed the current 10nm process is dead. He says that this is a good thing for the company as it was the right thing to do. Assuming that there is truth to this rumor, this means that only TSMC and Samsung are the gatekeepers of the next generation process. GLOBALFOUNDRIES has dropped out of the race. Samsung has begun production with their 7nm EUV process. TSMC announced that they expect their 7nm process to account for 20% of their revenue for 2019. There has been no official statement from Intel so far about the rumor.

For several years now SemiAccurate has been saying the the 10nm process as proposed by Intel would never be financially viable. Now we are hearing from trusted moles that the process is indeed dead and that is a good thing for Intel, if they had continued along their current path the disaster would have been untenable. Our moles are saying the deed has finally been done. This isn't to say the road to this point has been easy or straightforward, and the road ahead is even less solid.
I did read elsewhere that AMD killed Kaveri, Steamroller, and Excavator:

https://semiaccurate.com/2012/11/19/amd-kills-off-big-cores-kaveri-steamroller-and-excavator/

It must be true. AMD denied the history, but I guess "SemiAccurate stands by its reporting."
 
Top