Intel Stuck with 14nm Processors Till Holiday 2019

[dorsetknob]

there are so many official holidays here in 'murica. Pick one.

Inauguration Day

 

[efikkan]

So once again what are we talking about here as just saying 14,12,10,7nm says nothing.
Global Foundries 12nm processes is inferior to the above 14nm processes by all accounts save for the minimum feature size which is the magical 12nm which we obsess over.
Unless anyone here knows the ins and outs of Global Foundries 7nm* node then you can't compare and say who is ahead of what.

Yes, it's impressive how "everyone" knows that TSMC and Samsung "7nm" will be fantastic. Not only have all the previous iterations struggled a lot, even with volume shipments in Q1 2019, we still not know the shipping quantities and how large chips will get decent yields. Additionally people seem to swallow the PR material, forgetting that these are theoretical figures, and that the full potential will only be reached towards the end of the node life cycle.

Which actually means "Paper Launch December 2019 at the earliest." These probably will be laptop and low-power only like Broadwell.

They already have, it's called Cannon Lake.

There's nothing they CAN DO after they launch the 8/16 "i9" @ 5GHz. That's their last hurrah for 2 years.

New architectures take time to make, and they can't even really nail one down until they get 10nm working.

They already have their next architecture ready; Ice Lake. And despite all the terrible delays, it should allow Intel to tweak it prior to release more than they have for any architecture the past decade.

 

[HTC]

Dunno if this move won't actually be worse for Intel later on, on the 10nm node. I'm of the opinion that this is part of the current problem with 10nm already, besides the yield issues ofc.

Think about it: if they keep refining 14nm further, it will only be harder for 10nm to surpass it later on, in speed and / or efficiency and / or power usage with the same / more cores and with / without HT, unless they come up with some kind of "new Sandy Bridge" which can give them the possibility to leapfrog the difference between a super-refined process VS an early smaller process (it's not impossible, but it's unlikely).

Intel seems to be digging a bigger and bigger hole ...

 

[mcraygsx]

"10 nm process will arrive only by Holiday 2019"
Is that the 13th month for those of us not in the U.S.?

That is paper launch by Fall 2019. We might not see high performing processors in our desktops till 2020.

 

[efikkan]

Dunno if this move won't actually be worse for Intel later on, on the 10nm node. I'm of the opinion that this is part of the current problem with 10nm already, besides the yield issues ofc.

Think about it: if they keep refining 14nm further, it will only be harder for 10nm to surpass it later on, in speed and / or efficiency and / or power usage with the same / more cores and with / without HT, unless they come up with some kind of "new Sandy Bridge" which can give them the possibility to leapfrog the difference between a super-refined process VS an early smaller process (it's not impossible, but it's unlikely).

Intel seems to be digging a bigger and bigger hole ...

Well, there aren't much more potential to gain from 14nm, so any tweaks would be in the CPU design itself.

Even if Intel's 10nm isn't good enough for a full lineup one year from now, they still have one option; use 10 nm on select models, and 14 nm on the rest. They should prioritize 6-/8-core desktop and some mobile CPUs for 10nm, and with Intel's manufacturing volume they should be able to, even if the yields are so low they might have terrible margins on select products. Even if needed, Intel can take a loss on select models for a year. It's easy to forget that most of the shipped volume doesn't have to be on 10 nm immediately.

We should expect clocks to be lower on initial Ice Lake CPUs, but primarily on the boost side. Remember that Ice Lake is a major architecture unlike Zen2, so we should expect some IPC gains to offset the lower clock speed. Also remember that Zen2 will not reach the full gains of "7nm" by Q3 2019, AMD might get closer to Intel in core speed, but is extremely unlikely to surpass them.

 

[rtwjunkie]

Inauguration Day

I sure don’t get that day off.

Specifically, when the term “the holidays” is used in the U.S., it typically refers to the Christmas-New Year time period.

 

[ShurikN]

Specifically, when the term “the holidays” is used in the U.S., it typically refers to the Christmas-New Year time period.

Yeah but for example, Christmas in my country is on 7th of January. So the 2019 New year-Christmas holiday, by my mindset is in January of 2019 instead of December. It always takes me a while to reevaluate the calendar in my brain whenever I hear or read "Holiday" season.
It's one of the reasons I hate the term.

 

[rtwjunkie]

It's one of the reasons I hate the term.

Me too, because it really depends on where you use the term, and can be confusing in an international setting. I was just explaining it though, by my best guess.

 

[Durvelle27]

Intel may never get 10nm implemented

 

[HTC]

Well, there aren't much more potential to gain from 14nm, so any tweaks would be in the CPU design itself.

Even if Intel's 10nm isn't good enough for a full lineup one year from now, they still have one option; use 10 nm on select models, and 14 nm on the rest. They should prioritize 6-/8-core desktop and some mobile CPUs for 10nm, and with Intel's manufacturing volume they should be able to, even if the yields are so low they might have terrible margins on select products. Even if needed, Intel can take a loss on select models for a year. It's easy to forget that most of the shipped volume doesn't have to be on 10 nm immediately.

We should expect clocks to be lower on initial Ice Lake CPUs, but primarily on the boost side. Remember that Ice Lake is a major architecture unlike Zen2, so we should expect some IPC gains to offset the lower clock speed. Also remember that Zen2 will not reach the full gains of "7nm" by Q3 2019, AMD might get closer to Intel in core speed, but is extremely unlikely to surpass them.

According to this, from which the following pic is taken, you're wrong dude:

In the performance readiness part of the pic, you can clearly see that 14nm++ is ahead of even 10nm+ in performance, albeit slightly, and the performance increase over "regular" 14nm is quite substantial. If they end up refining 14nm++ into 14nm+++, it's possible that it can be close enough to dent the 10nm++, which is why i said:

Dunno if this move won't actually be worse for Intel later on, on the 10nm node.

Intel is in serious trouble here. On one hand, if they release 10nm as is, it will have lower clocks then the current refined 14nm and the lower yields make it costlier to produce so it's a lose - lose situation but, on the other hand, if they opt for an even more refined14nm process again, they'll be taking a serious chunk from the advantages of the newer 10nm by having this 14nm part perform as well if not slightly better, though @ a higher power envelope, so Intel will be in this exact situation later and that makes it also a lose - lose situation.

This is exacerbated by the arrival of the Zen architecture because, if AMD hadn't "rysen from the dead" (pun intended), Intel wouldn't be as pressured as it is now.

 

[efikkan]

According to this, from which the following pic is taken, you're wrong dude:

Well, there aren't much more potential to gain from 14nm, so any tweaks would be in the CPU design itself.

In the performance readiness part of the pic, you can clearly see that 14nm++ is ahead of even 10nm+ in performance, albeit slightly, …

I know that very well, but that's not what I meant at all, also you should have read the end of my post:

We should expect clocks to be lower on initial Ice Lake CPUs, but primarily on the boost side. Remember that Ice Lake is a major architecture unlike Zen2, so we should expect some IPC gains to offset the lower clock speed.

To get back to my first point, Intel can't squeeze much more out of 14nm++, so if they want more, it would have to be in the CPU design itself.
A good backup plan for 2018 would have been to backport Ice Lake to 14nm, but that's too late now.

Intel is in serious trouble here. On one hand, if they release 10nm as is, it will have lower clocks then the current refined 14nm and the lower yields make it costlier to produce so it's a lose…

Intel claims 10nm yields is "on track" for 2H 2019.

 

[Caring1]

Intel claims 10nm yields is "on track" for 2H 2019.

But it's been derailed before …..

 

[R0H1T]

Intel claims 10nm yields is "on track" for 2H 2019.
View attachment 104533

At this point it should be clear that whatever Intel's claiming is a lie, not that other corporations are much different but if you've been following Intel, they sound like you know who.

 

[efikkan]

But it's been derailed before …..

Sure, plans can always change, and unforeseen major events (like a flood affecting harddrive production).
But the yields are at least improving.

At this point it should be clear that whatever Intel's claiming is a lie, not that other corporations are much different but if you've been following Intel, they sound like you know who.

But we should clearly take everything AMD says as hard truth…

 

[StrayKAT]

I understand not liking them, but the amount of doubt on a company as old as Intel is just silly. You're treating them like amateur electricians who work at a flea market or something. I think they understand timing their products better than anyone.

 

[Kohl Baas]

I understand not liking them, but the amount of doubt on a company as old as Intel is just silly. You're treating them like amateur electricians who work at a flea market or something. I think they understand timing their products better than anyone.

AFAIK this isn't a matter of timing. They struggling to kickstart 10nm production on the necessary performance, cost and yield. If they can't find a solution to whatever problem they don't expected, somewhere, something will be lost. Well, if we're looking it realisticly some loss is already made because the previous deraile. All they can do is keep working on it and calculate what they want to loose.

 

[ShurikN]

I understand not liking them, but the amount of doubt on a company as old as Intel is just silly. You're treating them like amateur electricians who work at a flea market or something. I think they understand timing their products better than anyone.

While everything you said stands correct, 10nm was originally scheduled for 2015. It's 4 years late. You could argue they overestimated by a year and a half, tops... but 4?!
One would have assumed that a company like Intel with billions to pour into R&D could overcome the issue.

 

[ssdpro]

Intel is in serious trouble here.

It's a good thing Intel has an 80% market share and earned revenue of 62.8 billion last year. And 16.96 billion last quarter putting them on pace for 68 billion this year. I am not sure the bean counters even know what node process is as it doesn't seem to mean anything to $$$. People keep buying the best.

 

[R0H1T]

They already have, it's called Cannon Lake.

CNL is MIA, it has exactly 1 working SKU IIRC, in a Lenovo laptop & is so bad that the IGP is dysfunctional.
If we're talking about mass production, Intel scale, that's like saying the Tesla can mass produce model X just because they did fulfil their last minute promise of weekly model 3 deliveries.

Sure, plans can always change, and unforeseen major events (like a flood affecting harddrive production).
But the yields are at least improving.


But we should clearly take everything AMD says as hard truth…

This is the exact reason why I said ~

not that other corporations are much different

And yet here we are!

 

[HTC]

I know that very well, but that's not what I meant at all, also you should have read the end of my post:

To get back to my first point, Intel can't squeeze much more out of 14nm++, so if they want more, it would have to be in the CPU design itself.
A good backup plan for 2018 would have been to backport Ice Lake to 14nm, but that's too late now.


Intel claims 10nm yields is "on track" for 2H 2019.
View attachment 104533

Intel's 10nm is very very late, dude:



The above pic is taken from this: as can be seen in the pic's subtitle, Intel's 5nm was expected to arrive in 2019 @ the time of the article (2012), with 7nm in between 10nm and 5nm.

 

[efikkan]

Intel's 10nm is very very late, dude:

Please be more serious than this.
And how is this relevant? We all know it's delayed compared to the original plans, plans which have been revised many times and are no longer relevant.

Intel have been shipping very low volumes of 10 nm since April, and since they say yields are now on track for a volume launch 2H 2019, it means they are slowly improving.

 

[HTC]

Please be more serious than this.
And how is this relevant? We all know it's delayed compared to the original plans, plans which have been revised many times and are no longer relevant.

Intel have been shipping very low volumes of 10 nm since April, and since they say yields are now on track for a volume launch 2H 2019, it means they are slowly improving.

Each of these delays have cost money, and a lot of it: nearly 4 years is almost 2 "full architectures" worth of time. Intel won't go bankrupt obviously but that doesn't mean this doesn't severely affect them and i wouldn't be surprised if the money lost thus far from these delays, in total (directly and indirectly) surpassed the $20B already.

They could manage to sort out their 10nm problems even before 2H 2019 for all we know but Intel saying it won't reassure most due to the consecutive delays: if it were board manufacturers saying it, it could potentially have more credibility, @ this point.

 

[Rahmat Sofyan]

Hmmm, this is why AMD stock price getting higher ?

 

[HTC]

Hmmm, this is why AMD stock price getting higher ?

And that is relevant to this specific topic because ...?

 

[dalekdukesboy]

What. This is not even about AMD, its about process and fab development more than anything else. AMD is not even playing that game to begin with. And that 'choice' is starting to pay off for them now, while it did hurt them until recently.

Even so its not like Intel suddenly cannot compete when AMD puts out product on 7nm. Their 14nm isn't thát far off to make it obsolete. They are going to have to play the pricing game.

Hmmm pricing game...who had to play that one last and how did it work out for them? Not so well, just ask AMD selling their dirt cheap processors that couldn't touch x79 until present day, where now tables are turned and yeah Intel will live but this won't be fun for them, period.