Intel 7nm in 2021

[Taf the Ghost]

12 minutes ago, Master Disaster said:

I have a theory about Intels current situation, it's probably totally incorrect but meh, I'll share it anyway.

 

I'm not sure Intel are having trouble getting to 7nm, I think they've realised what they currently have is no match for what Zen will become and instead of focusing on only a node shrink, I think they're taking the time to do a full architecture redesign based around 7nm. Like AMD did when moving from FX to Zen, throw everything away and start again.

 

While they're doing it they can continue to improve the current architecture meaning unlike AMD they have competitive product on shelves and when they're ready they can drop the new line and retake the performance crown.

 

I'm struggling to believe that Intel are having so many problems getting to a node size others are already talking about surpassing, there's something else going on at blue camp.

They were supposed to have a brand new uArch in 2020 on Sapphire Rapids.  Scuttlebutt is that architecture is now called Granite Rapids and will follow Sapphire Rapids. They were going to split up the development of Server and Desktop cores quite a bit. (Still seems like they are doing that, but it'll be more of Desktop is a cut-down version of Server at the design level.) 

 

The issue for Intel is that Coffee Lake, Comet Lake and Rocket Lake were never supposed to exist, but we're going to be seeing those as Desktop products for the next 3 years. Intel's 7nm is itself still behind schedule, but it's hard to say how much of that is due to 10nm being a disaster or its own troubles. We do have it on pretty good understanding that Intel moved several Fabs from transitioning 10nm to just going right to 7nm. (This caused the whole "10nm is dead! Long live 10nm!" issue.)

 

Intel's Fabs have been a mess since 14nm launched. If they don't get 7nm well sorted, they're in a world of trouble.

[porina]

I can't keep track of all the lakes, but Intel are making the first meaningful architecture change since Skylake with Sunny Cove going into... whatever the 10nm mobile CPUs are called. From a technology perspective, that is more interesting to me than the process, although of course both are important in different ways.

[cj09beira]

14 hours ago, _Syn_ said:

Well hopefully they can still bring competition to AMD in those 2 years, it helps no one to favor one company over the other, I want both of them to innovate

thats not really the best thing, amd is small, really small and they need a good cash injection, having them dominate for a few years is actually the best thing long term, so that amd can keep up as R&D costs keep increasing.

for now their biggest achievement is paying most of their debt, by next year it should be gone, some very good years after that will be important to allow them to be competitive as we approach the last stretch of silicon manufacturing 

[.Apex.]

1 minute ago, cj09beira said:

thats not really the best thing, amd is small, really small and they need a good cash injection, having them dominate for a few years is actually the best thing long term, so that amd can keep up as R&D costs keep increasing.

for now their biggest achievement is paying most of their debt, by next year it should be gone, some very good years after that will be important to allow them to be competitive as we approach the last stretch of silicon manufacturing 

AMD is small? what? AMD is expanding rapidly, I doubt competition will affect them at this point, they're providing hardware for the PS5 and Xbox which will sell like hot cakes, their SOC designs, and now they're expanding to laptops, and there's EPYC for server hardware, and Navi is soon, etc etc

AMD is in a really good standing right now, this IS the right time for competition

[mr moose]

55 minutes ago, porina said:

I can't keep track of all the lakes, but Intel are making the first meaningful architecture change since Skylake with Sunny Cove going into... whatever the 10nm mobile CPUs are called. From a technology perspective, that is more interesting to me than the process, although of course both are important in different ways.

I'm more interested in how the business operates around these conditions.  While everyone is fixated on node advantages, Anyone watching their business model would have seen that it's just been the icing on the cake,  they seem to be doing just fine and I wait with a baited breath to see how the market changes and what happens next when TMSC goes into full production.   

 

My prediction is Intel can squeeze another few ++ out of 14nm, expand upon the current 10nm and might even bring 7nm to consumers earlier than 2021.  This is not based on watching the tech, but watching the business.

[GoldenLag]

1 hour ago, _Syn_ said:

AMD is small? what?

Yes, AMD is small. Like quite small when it comes to both income and spending. 

 

Nvidia is larger than them. And that is a company that practically only makes GPUs and some smaller arm chips.

[Silentprototipe]

And then in 2021 we will get 14nm+++++. Maybe 10nm by 2024

[samcool55]

Does this mean the AMD launch is close?

It might be just me but Intel has been constantly for the last few years trying to steal the spotlight from AMD by saying something "interesting" a few days/weeks before AMD does something.

 

I think the 28-core intel announcement and the threadripper 2 announcement are a perfect example of this.

[porina]

1 hour ago, samcool55 said:

Does this mean the AMD launch is close?

AMD have a keynote at Computex at end of month which is where people think they will announce Zen 2 for consumers.

[samcool55]

2 minutes ago, porina said:

AMD have a keynote at Computex at end of month which is where people think they will announce Zen 2 for consumers.

Well that might explain it. Let's hope this "big" news from intel does indeed mean big news from AMD as well soon.

As far as rumors go zen 2 does sound good so let's wait and see how that goes I guess.

[pas008]

general purpose gpu? love how they call it that

7nm is independently worked on

guess more news here kinda regurgitated though

https://www.engadget.com/2019/05/08/intel-ships-volume-10nm-in-june/?yptr=yahoo

[JustWantTech]

I heard that Intel 10nm = AMD 7nm. Is that true?

[Guest]

Quote

What is relatively surprising is that the company intends to start production of chips using its 7 nm process already in 2021.

[GoldenLag]

42 minutes ago, JustWantTech said:

I heard that Intel 10nm = AMD 7nm. Is that true?

That might be the case with the target performance Intel Aimed for. Though their yields have been poor. Also its TSMC 7nm. AMD/glofo arent in the race to 7nm. 

[Stefan Payne]

On 5/9/2019 at 1:01 PM, _Syn_ said:

AMD is small? what?

https://www.intc.com/investor-relations/investor-education-and-news/investor-news/press-release-details/2019/Intel-Reports-First-Quarter-2019-Financial-Results/default.aspx

http://ir.amd.com/news-releases/news-release-details/amd-reports-first-quarter-2019-financial-results

 

1.27 Billion vs. 16 Billion Revenue...

On 5/9/2019 at 1:01 PM, _Syn_ said:

AMD is in a really good standing right now, this IS the right time for competition

So you're saying that Only Intel should stay on top and there shouldn't be any competition??

 

38 minutes ago, JustWantTech said:

I heard that Intel 10nm = AMD 7nm. Is that true?

No, because AMD doesn't have any manufacturing capabilitys these days.

Its tSMC but its possible.

The Problem:

When Intel gets 7nm ready, TSMC might have 5nm ready.

[.Apex.]

34 minutes ago, Stefan Payne said:

https://www.intc.com/investor-relations/investor-education-and-news/investor-news/press-release-details/2019/Intel-Reports-First-Quarter-2019-Financial-Results/default.aspx

http://ir.amd.com/news-releases/news-release-details/amd-reports-first-quarter-2019-financial-results

 

1.27 Billion vs. 16 Billion Revenue... 

Smaller for sure, but they're not a small company
 

34 minutes ago, Stefan Payne said:

So you're saying that Only Intel should stay on top and there shouldn't be any competition??

I never said that... what? I said competition, either one of them can race to the top, but neither one of them should stay at the top forever, can't believe i have to explain this

[cj09beira]

2 hours ago, JustWantTech said:

I heard that Intel 10nm = AMD 7nm. Is that true?

on paper at least for density yes they are very similar, though 7nm is here and 10nm is nowhere to be found, also the details we have are probably wrong as they most likely had to reduce the density targets to get yields high enough

[Stefan Payne]

12 hours ago, _Syn_ said:

Smaller for sure, but they're not a small company

Yes, they are small.

So small that they can barely compete and also there is the "Mindshare" Problem.

 

12 hours ago, _Syn_ said:

I never said that... what? I said competition, either one of them can race to the top, but neither one of them should stay at the top forever, can't believe i have to explain this

Yes, you did!
You said that Intel should compete with AMD, who are 1/10th of the size of Intel and also do GPU as well, wich Intel doesn't really do right now.

Right now they are competing with Intel. And you said that Intel should compete with them, wich implies that Intel should "destroy AMD".

 

Right now we need AMD to make some money, to get Market- and Mindshare, to increase R&D Budget, so that they can put out GPUs that give Jensen a Tantrum any day because they can get GPU Market Share back. That might be with NAVI but also with others.

BUT right now, they are a drop in the Bucket compared with Intel. You seen the Data yourself!
Intel puts what AMD makes each Quater in R&D...

 

So yeah, you're wrong, your statement was bad, because both Companys need to have similar Market Share and make amounts of money that allow them good R&D to really compete.

[LAwLz]

Just to remind everyone that Intel's #nm is not comparable to for example TSMC's #nm.

The two are very different, just like you can't look at a phone and go "wow phone has an 8 core processor so therefore it's as fast as the Ryzen 2700 which is also an 8 core".

 

 

Here is a projected comparison of Intel 7nm vs TSMC 5nm.

Spoiler

 

As you can see. Intel's 7nm lithography is estimated to be about 37% higher density than the 5nm from TSMC.

Source: https://www.semiwiki.com/forum/content/7544-7nm-5nm-3nm-logic-current-projected-processes.html

 

We can also see on that page that Intel's 10nm is denser than TSMC's 7nm by about 10%.

But it's also worth keeping in mind that density is not the only important factor.

[Unixsystem]

9 hours ago, LAwLz said:

Just to remind everyone that Intel's #nm is not comparable to for example TSMC's #nm.

The two are very different, just like you can't look at a phone and go "wow phone has an 8 core processor so therefore it's as fast as the Ryzen 2700 which is also an 8 core".

 

 

Here is a projected comparison of Intel 7nm vs TSMC 5nm.

  Hide contents

 

As you can see. Intel's 7nm lithography is estimated to be about 37% higher density than the 5nm from TSMC.

Source: https://www.semiwiki.com/forum/content/7544-7nm-5nm-3nm-logic-current-projected-processes.html

 

We can also see on that page that Intel's 10nm is denser than TSMC's 7nm by about 10%.

But it's also worth keeping in mind that density is not the only important factor.

While all of that is true on paper, the problem I have with everyone constantly bringing this up is that Intel 10nm is still AWOL for all practical purposes, whereas TSMC has been shipping 7nm for months (and that's not even accounting for the fact that "Intel 10nm" isn't even really Intel 10nm anymore since they effectively loosened a lot of their design constraints a year ago or so. 

 

When it comes to Intel 7nm VS TSMC 5nm, obviously everything is just on paper, but at this point if I needed to put money on one getting delayed I'm betting on Intel. While it seems like the sub-12nm nodes have been a real bitch for everyone, Intel still has nothing to show for it and hasn't shipped a new node since Broadwell. Even if 10nm's failure doesn't directly impact 7nm's development, I have to imagine that them jumping in the deep end with an even smaller node when they haven't even made the easier nodes work isn't going to bode super well for them. 

 

Tl;dr: Even if Intel's nodes are 10x better than TSMC's on paper, TSMC's nodes actually exist. 

[Juanitology]

On 5/8/2019 at 4:57 PM, Bouzoo said:

"World class packaging technology complements process leadership"

Not the kind of packaging you're thinking of.

[PlayStation 2]

23 hours ago, Stefan Payne said:

Yes, you did!
You said that Intel should compete with AMD, who are 1/10th of the size of Intel and also do GPU as well, wich Intel doesn't really do right now.

Right now they are competing with Intel. And you said that Intel should compete with them, wich implies that Intel should "destroy AMD".

 

Right now we need AMD to make some money, to get Market- and Mindshare, to increase R&D Budget, so that they can put out GPUs that give Jensen a Tantrum any day because they can get GPU Market Share back. That might be with NAVI but also with others.

BUT right now, they are a drop in the Bucket compared with Intel. You seen the Data yourself!
Intel puts what AMD makes each Quater in R&D...

 

So yeah, you're wrong, your statement was bad, because both Companys need to have similar Market Share and make amounts of money that allow them good R&D to really compete.

OP probably muddled his words a little (I had to figure out what he probably meant) but I highly doubt he meant that only Intel should stay at that top. I think his implication was that both companies need to stay competitive to make money rather than one company dominate the market.

[mr moose]

On 5/9/2019 at 9:01 PM, _Syn_ said:

AMD is in a really good standing right now, this IS the right time for competition

It's always the right time for competition.  Good honest competition.  Keeps them working hard on better tech and prices competitive.

 

 

To the rest of the thread:

 

I don't know why people are hung up on market share though,  having market share doesn't give you better products, having better products gives you market share.    And stop talking about R+D  budgets.  AMD's R+D budget is the same today as it was in 2006 and the only period it slumped down in was the 2 years prior to Ryzen release and it has been slowly rising since.  So trying to use R+D budget as a predictor for future product performance is littered with errors.  For one money is not always spent wisely and second even if it is spent wisely there is still no guarantee the competition didn't do it better.

 

 

[Taf the Ghost]

1 minute ago, mr moose said:

It's always the right time for competition.  Good honest competition.  Keeps them working hard on better tech and prices competitive.

 

 

To the rest of the thread:

 

I don't know why people are hung up on market share though,  having market share doesn't give you better products, having better products gives you market share.    And stop talking about R+D  budgets.  AMD's R+D budget is the same today as it was in 2006 and the only period it slumped down in was the 2 years prior to Ryzen release and it has been slowly rising since.  So trying to use R+D budget as a predictor for future product performance is littered with errors.  For one money is not always spent wisely and second even if it is spent wisely there is still no guarantee the competition didn't do it better.

 

 

No one talks about the wasted billions in R&D Intel spends that never goes anywhere.

[mr moose]

1 minute ago, Taf the Ghost said:

No one talks about the wasted billions in R&D Intel spends that never goes anywhere.

 

No, they just assume the reason Intel's products sell is because every single cent went into precisionally executed Research and development.