[DigiTimes] TSMC 10 nm trial production in 2015, mass production in 2016

[Arachnotronic]

Is it just a jockeying game for investors?

Yes.

 

[uclaLabrat]

Heaviside step function

What most people invoke as a "quantum leap" is really more appropriately described as a Heaviside step function whenever the individual invoking the phrase "quantum leap" is doing so to imply a gap or step-up in whatever is being discussed versus the less relevant equivocation of the physics entailed in a quantum transition.

If you are looking for a specific example, TSMC doubled their entire R&D team (headcount and resources) for 10nm development.



2 years ago I would have taken that bet. But I'm not seeing Intel running the same way under BK that it did under Otellini. I think we are going to see 10nm run head-to-head, and then 7nm, etc., even if by virtue of marketing manipulation of the node label. If Intel gets too far ahead, again, then the foundries will just pull another 16/14nm out (claim a new node exists where none was before) and they'll be back to "node parity".

As for those who will rush to internalizing that as an erosion in Intel's lead...true but that is Intel's problem. Intel has no issue hyping their technology prowess for the benefit of analyst and shareholder alike. I'm confident they won't sit idly by and let their shareholders and analysts down. They'll find a marketing way to ensure the market is fully and properly educated on why their 10nm (or 7, or 5) is superior to the competition's.



Ha ha, 10,000pm We stuck fast to the micron node label well near the 0.1um node before finally switching vernacular and referring to them in nanometer units. Even the early days of 90nm development it was still referred to as 0.09um. I doubt we'll ever see nm node labeling give way to pm node labels, the numbers are just to big to make for good marketing.

It will be nm all the way down to 0.1nm (equivalence, of course, the same as we use for gate oxide scaling which is given an EOT rating in Angtroms).

EUV, if it happens in production at 10nm, will likely be limited to just a single litho step at the very most critical mask in the entire flow, with all the remaining less-critical (but still paramount) masks being done in 2, 3 or 4 multi-patterning. But it is a necessary "first" for EUV to gain the production environment time it needs to work out all the gatekeeping issues before full production rollout at 7nm can happen.



Intel could, but nothing is keeping TSMC and Samsung from deciding they are skipping 10nm and are going to go straight to 7nm...the key being their 7nm would look surprising similar to what they were previously calling their 10nm node :hmm: :sneaky: (kinda like how 16FF/14nm look an awful lot like what they should have delivered with their 20nm nodes in the first place...)

Not to pick nits, but do you really think they can make transistor features one Si atom thick? I'm skeptical, to say the least.

 

[Idontcare]

Is it just a jockeying game for investors?

Yes.

A very small percentage of a semi-company's employees are engineers who would know the technical merits of their company's process node versus a competitors.

Even within the company there can exist a sizable portion of the workforce who believe their own marketing, so it does have an employee morale effect.

In terms of marketing there is a huge effect that goes into convincing customers that the latest and greatest is the best. (and no, we end-users are not the customer, all these foundries and IDMs sell to customers who in turn sell us end-users an integrated product)

This sales and marketing hook will only continue to grow stronger as time goes on.

And lastly, as Arachnotronic spoke to, is the investment community. Shareholders and investors who buy for a myriad number of reasons, including ratings placed on stocks by analysts. All these people are swayed by the perception of the R&D value that is coming forth from any given company's R&D organization, as well as the production capabilities once that tech is ramped up.

In short, it very much is used to fool anyone and everyone that can be fooled, which basically leaves just a handful of process and design engineers who have the educational and experiential background to know an apple from an orange when it comes to process nodes

 

[Xpage]

Heaviside step function

What most people invoke as a "quantum leap" is really more appropriately described as a Heaviside step function whenever the individual invoking the phrase "quantum leap" is doing so to imply a gap or step-up in whatever is being discussed versus the less relevant equivocation of the physics entailed in a quantum transition.

IDC confirms 10nm Quantum computers from intel are coming. Time to start a rumor site.

 

[Idontcare]

IDC confirms 10nm Quantum computers from intel are coming. Time to start a rumor site.

D:

Please don't

 

[NTMBK]

IDC confirms 10nm Quantum computers from intel are coming. Time to start a rumor site.

I give it a week before this is a WCCFtech story.

 

[witeken]

2 years ago I would have taken that bet. But I'm not seeing Intel running the same way under BK that it did under Otellini. I think we are going to see 10nm run head-to-head, and then 7nm, etc., even if by virtue of marketing manipulation of the node label. If Intel gets too far ahead, again, then the foundries will just pull another 16/14nm out (claim a new node exists where none was before) and they'll be back to "node parity".

I don't think Krzanich changed anything in the node development. He's an engineer, and there are also other people like Holt:

“[...] When we began looking at the increasing wafer cost we had a somewhat substantial debate as to what should be that point for 14nm [on the cost per transistor curve]. And when confronted with a substantial increase in wafer cost and a historical trend of density scaling... we had considered whether we would have to begin to bend this curve [towards the end of Moore's Law]. And there was in fact a substantial opinion that it was now time to start bending the curve. But we stepped back and [...] we got a point substantially better than what we had asked for. We are in fact accelerating Moore's Law.” --William Holt, Intel, IM’13

Intel's 10nm will go into production in Q4, if there are no problems. Early work on TSMC's 10nm will be started at the end of the year. Volume production 2017, which is the same year 7nm is planned to go into production (usual disclaimer), so I think 10nm will compete less than or about a year with other foundries' 10nm.

Don't forget that 14nm is here in big volumes, with no sign of other companies' 14/16 node.

Ha ha, 10,000pm We stuck fast to the micron node label well near the 0.1um node before finally switching vernacular and referring to them in nanometer units. Even the early days of 90nm development it was still referred to as 0.09um. I doubt we'll ever see nm node labeling give way to pm node labels, the numbers are just to big to make for good marketing.

It will be nm all the way down to 0.1nm (equivalence, of course, the same as we use for gate oxide scaling which is given an EOT rating in Angtroms).

People don't like to change their habits. I once read an article by Chipworks about the history of Intel's nodes up to 45nm, but they kept using the [d...] microns, and not even with µm: "0.045 ‘m gate length", etc.

http://www.chipworks.com/en/technic.../blog/the-path-from-pentium-to-penryn-part-2/

 

[Fjodor2001]

Intel's 10nm will go into production in Q4, if there are no problems. Early work on TSMC's 10nm will be started at the end of the year. Volume production 2017, which is the same year 7nm is planned to go into production (usual disclaimer), so I think 10nm will compete less than or about a year with other foundries' 10nm.

Sources for all of those claims, please?

 

[AtenRa]

Don't forget that 14nm is here in big volumes, with no sign of other companies' 14/16 node.

We are in late January 2015 and only a handful of Core M devices are in the market. 5th Core Broadwell Laptop SKUs are not even been reviewed yet. I really dont see where the hell did you see big volumes of 14nm.


ps: remember this one ?? :whiste:

 

[Idontcare]

I don't think Krzanich changed anything in the node development. He's an engineer, and there are also other people like Holt:

Don't take this the wrong way, but do you know any Intel process dev engineers? Meaning have you had a chance to talk with one recently?

If you are solely relying on what Intel puts out in their PR then you will always be just as far behind the info curve as Intel wants their competition to be.

Think about it. Why has Intel privileged you or anyone else in the public domain with possessing information regarding anything they are or are not doing at 10nm?

The Intel press releases only make sense if you have the context with which to parse them. OEMs have that context by virtue of product pipeline schedules, which are NDA of course. And suppliers have access to the context by virtue of them working hand-in-hand with Intel to make the timeline happen in the first place. But who in the public domain has the context?

In Intel's ideal world no one does, because if you do then that means their competitor's might possibly as well. And the whole point of Intel's "keep quiet about 10nm" is to keep everyone in the public domain in the dark.

So you gotta find yourself someone who is in the industry who can give you context to the Intel PR, otherwise you are going to be intentionally lead to having the wrong conclusions. (because Intel would very much like to lead their competitor's to arriving at the wrong conclusions)

Intel's 10nm will go into production in Q4, if there are no problems. Early work on TSMC's 10nm will be started at the end of the year. Volume production 2017, which is the same year 7nm is planned to go into production (usual disclaimer), so I think 10nm will compete less than or about a year with other foundries' 10nm.

Don't forget that 14nm is here in big volumes, with no sign of other companies' 14/16 node.

Your Intel timeline seems fantastically optimistic.

TSMC garners >21% of their revenue from 20nm. Do you think >21% of Intel's revenue is from 14nm? 20nm is here in big volumes, they just helped Apple bag $18B in quarterly profits. What did 14nm do for Intel in Q4?

Personally I haven't laid my eyes on a single Intel 14nm processor. Have only seen a few integrated mobile devices containing them.

Newegg doesn't even list 14nm as an option when you use powersearch. Where are you getting the impression or data to support an argument that 14nm is here in big volumes? (serious question, did they discuss it in their recent earnings CC?)

 

[Magic Carpet]

@ Idontcare

Easily the best post in this whole thread :thumbsup:

 

[Arachnotronic]

We are in late January 2015 and only a handful of Core M devices are in the market. 5th Core Broadwell Laptop SKUs are not even been reviewed yet. I really dont see where the hell did you see big volumes of 14nm.


ps: remember this one ?? :whiste:

I'm typing this post on a Dell XPS 13 with a Broadwell Core i5.

 

[Fjodor2001]

I'm typing this post on a Dell XPS 13 with a Broadwell Core i5.

Congratulations, you're in possession of one of the handful few devices available. :awe:

 

[EightySix Four]

Don't take this the wrong way, but do you know any Intel process dev engineers? Meaning have you had a chance to talk with one recently?

If you are solely relying on what Intel puts out in their PR then you will always be just as far behind the info curve as Intel wants their competition to be.

Think about it. Why has Intel privileged you or anyone else in the public domain with possessing information regarding anything they are or are not doing at 10nm?
*SNIP*

Off Topic: When are you going to start a podcast?

 

[NTMBK]

I'm typing this post on a Dell XPS 13 with a Broadwell Core i5.

Okay, we found where half of the shipping chips went, who got the other one :awe:

(Seriously though, that looks like a sweet laptop. Would you recommend it?)

 

[Arachnotronic]

(Seriously though, that looks like a sweet laptop. Would you recommend it?)

Yes. Its touch pad isn't as nice as an Apple touchpad,but far better than most of the touchpads I've used on Windows machines. The display is great and the system itself feels pretty snappy. Keyboard is also quite good.

I personally recommend avoiding the high res model and going with the non-touch 1080P model. Also, probably no reason to pay extra for a Core i7 -- the Core i5 feels plenty quick.

All in all, I'm pretty happy with my purchase...well, except for the fact that I ordered it on day 1 but didn't get it until well after it was available locally at a Microsoft store. Don't order it directly from Dell if you can get it at retail...they take forever to ship.

 

[NTMBK]

Yes. Its touch pad isn't as nice as an Apple touchpad,but far better than most of the touchpads I've used on Windows machines. The display is great and the system itself feels pretty snappy. Keyboard is also quite good.

I personally recommend avoiding the high res model and going with the non-touch 1080P model. Also, probably no reason to pay extra for a Core i7 -- the Core i5 feels plenty quick.

All in all, I'm pretty happy with my purchase...well, except for the fact that I ordered it on day 1 but didn't get it until well after it was available locally at a Microsoft store. Don't order it directly from Dell if you can get it at retail...they take forever to ship.

Well that's annoying- in the UK they are only selling the hi-DPI and i7 model. :\ Ah well, we won't be replacing laptops for a while anyway. Thanks!

 

[AtenRa]

newegg will have Broadwell laptops only from 2-6-2015.

http://www.newegg.com/Product/Produc...40%20600551516

 

[Arachnotronic]

Newegg doesn't even list 14nm as an option when you use powersearch. Where are you getting the impression or data to support an argument that 14nm is here in big volumes? (serious question, did they discuss it in their recent earnings CC?)

re: 14nm volumes. Intel said at its 2014 investor meeting that it is shipping "millions" of Broadwell-U chips. Additionally, Intel has publicly stated that the transition from Haswell-U to Broadwell-U will be the fastest mobile transition in Intel's history (whatever that means).

I think you're going to see sizable quantities of 14nm product in the market by the end of Q1, and I think that by the end of the year the vast majority of Intel's product lines will have transitioned to 14nm.

 

[Arachnotronic]

Well that's annoying- in the UK they are only selling the hi-DPI and i7 model. :\ Ah well, we won't be replacing laptops for a while anyway. Thanks!

You're welcome! Yeah, for some reason the Windows PC vendors keep trying to push super hi-DPI models, but it has a negative impact on battery life and frankly Windows DPI scaling still isn't great. I much prefer 1080P.

 

[CHADBOGA]

Yes. Its touch pad isn't as nice as an Apple touchpad,but far better than most of the touchpads I've used on Windows machines. The display is great and the system itself feels pretty snappy. Keyboard is also quite good.

I personally recommend avoiding the high res model and going with the non-touch 1080P model. Also, probably no reason to pay extra for a Core i7 -- the Core i5 feels plenty quick.

All in all, I'm pretty happy with my purchase...well, except for the fact that I ordered it on day 1 but didn't get it until well after it was available locally at a Microsoft store. Don't order it directly from Dell if you can get it at retail...they take forever to ship.

What kind of battery life are you getting out of it?

 

[Fjodor2001]

What kind of battery life are you getting out of it?

Yeah, I'd like to know that too. Also, is it quiet in general? Does the fan always remain shut off during low performance work, such as web surfing / emailing / watching movies?

I'm considering this or a similar one myself, that's why I wonder. If it's too noisy, I'll optionally consider a Broadwell M Core Y based if it's not painfully slow.

 

[witeken]

Not to pick nits, but do you really think they can make transistor features one Si atom thick? I'm skeptical, to say the least.

Node names do not represent actual features sizes. And besides, discussing about how a 0.1nm node transistor might look like, is completely meaningless since even the smartest people like Bohr don't even know completely yet how a 5nm one will look like.

 

[witeken]

In short, it very much is used to fool anyone and everyone that can be fooled, which basically leaves just a handful of process and design engineers who have the educational and experiential background to know an apple from an orange when it comes to process nodes

How about forum people?

 

[witeken]

Sources for all of those claims, please?

10nm production: IM'13
TSMC claims: earnings call of some days ago.
7nm: extrapolation from 2 year Tick-Tock cadence and slides from Intel (but usual disclaimer that actual dates may vary materially since these are forward looking statements...)