TSMC catching up with Intel on leading edge process nodes

[witeken]

But the ambitions are high set. Lets just hope the economy doesn't go down before. Will be interesting to see after smartphone companies Q1 results.

Yeah, we all know how well TSMC executes its roadmaps. I will believe that they can execute when I see the product in market, not before they prove it.

16FF: November 2013

 

[Arachnotronic]

In the truth Intel's process lead did not change after the arrive of FinFets into the market. Intel's lead appears to be increasing, tho.
Also DK says Intel will get to QWFETs(the next big technique in transistor architecture, will be added to SiGe, HK/MG and FF to increase performance and lower power).



Is this. From what i know.

Idontcare said on these forums before he was chased off that 10nm is FinFET, no QWFET or any of the fancy stuff that Kanter predicted.

 

[Fjodor2001]

Will be interesting to see when TSMC can deliver products. Or if the dates needs the usual +2 years attached.

Yes, we all know you're claims regarding this:

It seems at 10nm that Intel may be 4-5 years ahead. Assuming that 10nm ever materialize for the foundries.

We're still waiting for your sources to back that up though. Got any?

 

[raghu78]

Yes, we all know you're claims regarding this:

We're still waiting for your sources to back that up though. Got any?

well said. he is as good as it gets when it comes to idolizing Intel and trashing Intel's competitors. TSMC was very confident of maintaining a 70% foundry share of FINFET in 2016 and a large share at 10nm from the start. They also said they don't intend to lose that 10nm share (a hint at Apple A9). My guess is TSMC have 100% of the A10/A10X business and A11/A11X business all tied up. Apple is driving the early ramp of these processes. I would not be surprised to see a 10nm A10X in Q2 2017 and a 7nm A11X in Q2 2018. Apple's A9X is already quite large at 147 sq mm and would benefit hugely from a 10nm shrink for the next gen A10X.

 

[ShintaiDK]

Yes, we all know you're claims regarding this:



We're still waiting for your sources to back that up though. Got any?

You have been told multiple times but you keep asking in your furious flaming attempts.

well said. he is as good as it gets when it comes to idolizing Intel and trashing Intel's competitors. TSMC was very confident of maintaining a 70% foundry share of FINFET in 2016 and a large share at 10nm from the start. They also said they don't intend to lose that 10nm share (a hint at Apple A9). My guess is TSMC have 100% of the A10/A10X business and A11/A11X business all tied up. Apple is driving the early ramp of these processes. I would not be surprised to see a 10nm A11X in Q2 2017 and a 7nm A11X in Q2 2018. Apple's A9X is already quite large at 147 sq mm and would benefit hugely from a 10nm shrink for the next gen A10X.

And that comes from the 300/Fiji 40% market share guy?

You should just have dropped the first line.

 

[NTMBK]

Yes, we all know you're claims regarding this:



We're still waiting for your sources to back that up though. Got any?

Good god man, get over this weird obsession with Shintai. Digging up posts from a year and a half ago to make some petty point? Jeeze.

 

[mikk]

I'm suprised that most people assume that all TSMC claims will be transformed into reality, but the history tells us the opposite. Don't they think delays may happen to TSMC? Also most people are not aware that TSMC Roadmaps refer to mass production ready while Intel Roadmaps are usually referring to product launch. And last I have to say initial 10nm/7nm products may be limited to small SOC chips for a longer time. We are still waiting for dedicated 16nm GPUs from Nvidia/AMD.

 

[raghu78]

And that comes from the 300/Fiji 40% market share guy?

I was optimistic about Fiji and R9 300 gaining market share and I was wrong as they were nothing more than a scaled Hawaii GPU with new mem tech and rebrands. I can accept a mistake unlike you who keep making some ridiculous statements and don't ever have the humility to accept your mistakes (and stop making them in future) when the reality proves you horribly wrong.

You should just have dropped the first line.

In that case you should stop posting because we are tired of keeping track of how many times you are wrong.

 

[frozentundra123456]

Do you really think it is your decision who can and cannot post? If you dont like what he posts, you are free to not read them or use the ignore button.

 

[Arachnotronic]

well said. he is as good as it gets when it comes to idolizing Intel and trashing Intel's competitors. TSMC was very confident of maintaining a 70% foundry share of FINFET in 2016 and a large share at 10nm from the start. They also said they don't intend to lose that 10nm share (a hint at Apple A9). My guess is TSMC have 100% of the A10/A10X business and A11/A11X business all tied up. Apple is driving the early ramp of these processes. I would not be surprised to see a 10nm A11X in Q2 2017 and a 7nm A11X in Q2 2018. Apple's A9X is already quite large at 147 sq mm and would benefit hugely from a 10nm shrink for the next gen A10X.

Timing too early on A12X here.

If mass production starts in 1H 2018 as indicated there's two cases.

Case 1: 1Q '18 production start

Cycle time for 16FF+ was around 1.5qtrs, so given the more intense litho I'd call 7nm cycle time at maybe more like 2qtrs. So wafers in January means wafers delivered to Apple around June, in time for a device that launches early in Q3'18.

Case 2: 2Q '18 production start

Start in April, this means wafers delivered just in the nick of time for a September unveiling/October device availability.

I think we won't see 7nm until the iPhone 8.

 

[superstition]

I was optimistic about Fiji and R9 300 gaining market share and I was wrong as they were nothing more than a scaled Hawaii GPU with new mem tech and rebrands.

Fiji is quite different from Hawaii which is why, for instance, it has much better performance per watt.

As for Arachnotronic's point about AMD staying on 14nm I think it has merit since every node shrink increases the number of design rules and thus the complexity/time/expense of putting a chip onto it. However, AMD may feel that staying on a node that's significantly weaker than Intel's isn't an option so they'll push forward Zen+ to a smaller one.

 

[Fjodor2001]

Good god man, get over this weird obsession with Shintai. Digging up posts from a year and a half ago to make some petty point? Jeeze.

The problem is he still believes his claims are correct, and refuse to admit they were completely inaccurate to begin with. And this is not a one time occurrence, he does it over and over again on the forum. I.e. make dead certain controversial fanboy claims, do not back it with sources, then when proven wrong refuse to admit he was wrong and still sticks with the original claim. That's what's weird and an obsession. So I think he really needs to be held accountable for that.

 

[Fjodor2001]

You have been told multiple times but you keep asking in your furious flaming attempts.

Yes, I keep asking the same question because you never respond and provide any sources to back up your claims. Yet you refuse to admit your claims are and were incorrect.

So how does your post prove that Intel is 4-5 years ahead of the competition on 10 nm?

And do you still believe 10 nm will never even materialize at the foundries?

 

[BarkingGhostar]

Fanboys

 

[Fjodor2001]

Yeah, we all know how well TSMC executes its roadmaps. I will believe that they can execute when I see the product in market, not before they prove it.

16FF: November 2013

Intel's roadmap, reality caught up with them:

 

[frozentundra123456]

Fiji is quite different from Hawaii which is why, for instance, it has much better performance per watt.

As for Arachnotronic's point about AMD staying on 14nm I think it has merit since every node shrink increases the number of design rules and thus the complexity/time/expense of putting a chip onto it. However, AMD may feel that staying on a node that's significantly weaker than Intel's isn't an option so they'll push forward Zen+ to a smaller one.

They probably have felt that for the last several years, but it hasnt stopped them.

 

[jpiniero]

And do you still believe 10 nm will never even materialize at the foundries?

There's materalize and then there's actually shipping products. You have to remember that nobody is going to use "10" nm unless it's comparable or cheaper than "14"/"16" nm except for maybe Apple. So maybe the 2018 iPhone will use it but it'd be 2019 or 2020 before you saw any products from nVidia or AMD for instance.

 

[Fjodor2001]

There's materalize and then there's actually shipping products.

So do you think 10 nm will not ever "materialize" at the foundries, like ShintaiDK claimed?

You have to remember that nobody is going to use "10" nm unless it's comparable or cheaper than "14"/"16" nm except for maybe Apple.

Source? And how would this be different from other semiconductor manufacturers like Apple?

So Samsung would just skip 10 nm (e.g. for Galaxy series) despite being able to produce such chips? Why would they do that!? They would just let their 10 nm capable factories sit idle for a couple of years? Please explain how that makes sense.

So maybe the 2018 iPhone will use it but it'd be 2019 or 2020 before you saw any products from nVidia or AMD for instance.

Source? Or just speculation? Why don't you speculate about Intel not releasing 10 nm products in 2020 too? What makes Intel different compared to TSMC/Samsung/GF? Are you expecting the latter to just let their 10 nm capable factories sit idle for several years, not producing anything!?

 

[jpiniero]

So do you think 10 nm will not ever "materialize" at the foundries, like ShintaiDK claimed?

It's possible if the economy falls into a recession. The design costs at "10" presumably are going to be even more expensive at that lower node. If you don't have the demand to justify it, it won't be done.

Source? And how would this be different from other semiconductor manufacturers like Apple?

Apple has shown that they will pay for access to smaller nodes... within reason of course.

 

[cytg111]

So when you guys tire of cherry picking benchmarks you alternate to cherrypicking old marketing slides. Good on you, you need to mix that diet up to stay healthy!

 

[superstition]

They probably have felt that for the last several years, but it hasnt stopped them.

Which is where the revenue from Zen comes in, along with potential help from Samsung and such.

 

[raghu78]

Timing too early on A12X here.

If mass production starts in 1H 2018 as indicated there's two cases.

Case 1: 1Q '18 production start

Cycle time for 16FF+ was around 1.5qtrs, so given the more intense litho I'd call 7nm cycle time at maybe more like 2qtrs. So wafers in January means wafers delivered to Apple around June, in time for a device that launches early in Q3'18.

Case 2: 2Q '18 production start

Start in April, this means wafers delivered just in the nick of time for a September unveiling/October device availability.

I think we won't see 7nm until the iPhone 8.

Arachnotronic I am of the opinion that A9X will not be refreshed till Q2 2017. A9X will power the next gen iPad Air 3. Apple might introduce A10X as the first TSMC 10nm chip. A10X desperately needs the area shrink and there is a good chance it will get it first. iPhone 7s will get to 10nm in late Q3 2017. btw TSMC expects to begin volume production for 10nm in late 2016 and for 7nm in late 2017.

http://www.tsmc.com/uploadfile/ir/quarterly/2015/4T5Um/E/TSMC 4Q15 transcript.pdf

Page 5

"Our 10-nanometer technology development is on track. We are currently in intensive yield learning mode in our technology development. Our 256-megabit SRAM is yielding well. We expect to complete process and product qualification and begin customer product tape-outs this quarter."
btw TSMC has confirmed 10nm product tapeouts in Q1 2016 and production by late 2016.

http://www.fool.com/investing/gener...im-expecting-from-the-apple-inc-ipad-pro.aspx

So Q2 2017 looks do-able for A10X and a year later A11X at 7nm in Q2 .

 

[nismotigerwvu]

The chance of that happening is zero.

IBM just paid GF to take their fab business. If they had a disruptive litho technology like that they wouldn't have simply given it away.

Oh I never said it would. My point was that if any competitor got the jump on shipping a functioning product made from a next gen material Intel would enter a full blown panic. You can substitute any company you wish and whatever material you want.

 

[Arachnotronic]

Yeah, we all know how well TSMC executes its roadmaps. I will believe that they can execute when I see the product in market, not before they prove it.

16FF: November 2013

That was the risk production schedule, but as we know 16FF never made it into production. They needed 16FF+ to be competitive.

 

[III-V]

The actual differences between nodes shrinks over time, both in regards to cost and performance. Exponential increases and decreases never last forever.

Intel's been getting jumps here and there and keeping ahead of the pack by being first to market with new technologies. If they fail to continue that trend, the rest of the pack will inevitably catch up. They need to go to 450 mm already.

Intel's density advantage isn't really relevant if they can't keep their fab utilization (and yields) up. Likewise, TSMC and Samsung have to do the same if they want to keep up.

Regardless of Intel vs. Samsung vs. TSMC, the biggest industry trend will be the cannibalization of companies that couldn't keep up with their fab tech.