By: juanrga (nospam.delete@this.juanrga.com), August 27, 2014 3:10 am
Room: Moderated Discussions
David Kanter (dkanter.delete@this.realworldtech.com) on August 26, 2014 6:25 pm wrote:
> > Intel is not targeting "real" 14nm. Intel deviated from ITRS
> > rules years ago, what they call 14nm is "real" ~16nm.
> >
> > TSMC and Glofo/Samsung are playing by same rules than Intel now, just to avoid
> > marketing abuse from Intel. As Scott Thompson (former Intel fellow) said: "Intel's
> > 22nm node is really 26nm, so if Intel does new math, so will we."
> >
> > What part of Intel 14nm is only about half-node away from TSMC 16nm is not stil understood?
>
> Repeat after me: The name of the node (e.g., 32nm, 16nm, 14nm, etc.) is simply
> marketing. Until you understand that, you are frankly just wallowing in your
> own ignorance and making yourself look bad and annoying the rest of us.
>
> It doesn't matter whether you call it 14nm, 16nm, 15nm or a zebra. What matters is:
>
> 1. Contacted gate pitch
> 2. Minimum metal pitch
> 3. Ion vs. Ioff curves for NFETs and PFETs
> 4. Design rule restrictions
> 5. Metal performance
> 6. Unique features e.g., eDRAM, high density capacitors, etc.
> 7. Yield
> 8. Time to market for comparable products
> 9. SRAM density
>
> Fortunately, #1,2,3,6, & 9 are often disclosed at VLSI or IEDM.
>
> Historically Intel was ahead on CGP and SRAM density, but behind on metal pitch (because
> they optimized for RC delay instead of density). With 14nm, Intel will be ahead on
> density as well, because the foundries are all using 20nm metal back-ends.
>
> The difference between Intel and TSMC should be judged on the basis
> of 1-9, not on the basis of BS marketing slides you've seen.
>
> David
I find insulting the condescending tone of your post, specially when you are trying to say me what I said before in this forum. I will try to write it again because you somewhat lose my posts.
Once again: TSMC "16nm" is so marketing label as Intel "14nm" is. I will copy-paste:

Now, if my actual point about nodes had been discussed we would not be going in circles.
> > Intel is not targeting "real" 14nm. Intel deviated from ITRS
> > rules years ago, what they call 14nm is "real" ~16nm.
> >
> > TSMC and Glofo/Samsung are playing by same rules than Intel now, just to avoid
> > marketing abuse from Intel. As Scott Thompson (former Intel fellow) said: "Intel's
> > 22nm node is really 26nm, so if Intel does new math, so will we."
> >
> > What part of Intel 14nm is only about half-node away from TSMC 16nm is not stil understood?
>
> Repeat after me: The name of the node (e.g., 32nm, 16nm, 14nm, etc.) is simply
> marketing. Until you understand that, you are frankly just wallowing in your
> own ignorance and making yourself look bad and annoying the rest of us.
>
> It doesn't matter whether you call it 14nm, 16nm, 15nm or a zebra. What matters is:
>
> 1. Contacted gate pitch
> 2. Minimum metal pitch
> 3. Ion vs. Ioff curves for NFETs and PFETs
> 4. Design rule restrictions
> 5. Metal performance
> 6. Unique features e.g., eDRAM, high density capacitors, etc.
> 7. Yield
> 8. Time to market for comparable products
> 9. SRAM density
>
> Fortunately, #1,2,3,6, & 9 are often disclosed at VLSI or IEDM.
>
> Historically Intel was ahead on CGP and SRAM density, but behind on metal pitch (because
> they optimized for RC delay instead of density). With 14nm, Intel will be ahead on
> density as well, because the foundries are all using 20nm metal back-ends.
>
> The difference between Intel and TSMC should be judged on the basis
> of 1-9, not on the basis of BS marketing slides you've seen.
>
> David
I find insulting the condescending tone of your post, specially when you are trying to say me what I said before in this forum. I will try to write it again because you somewhat lose my posts.
Once again: TSMC "16nm" is so marketing label as Intel "14nm" is. I will copy-paste:
What TSMC/IBM/Samsung/GF said is:
We call these processes 14nm / 16nm because:
Intel deviated from the official ITRS process naming rules and Intel
marketing/shareholders always abuse this against us to claim they
are many, many years ahead.
Indeed, based on BEOL density Intel's processes should have been called
differently according to the official ITRS naming standards:
Intel's 65nm should have been called 80nm instead.
Intel's 45nm should have been called 55nm instead.
Intel's 32nm should have been called 39nm instead.
You can see this in the graph below:
-The blue squares show the official metal 1 half pitches according to the ITRS.
-The purple circles and squares are the actual BEOL metal 1 half pitches of Intel's processes.
You can see that going from 90nm to 65nm Intel improved the BEOL MT1
by only 5%. They did exactly the same then as TSMC/IBM/Samsung/GF
are doing now.
It is due to the ignorance of the general public, which has no knowledge
of the official process naming rules, that TSMC/IBM/Samsung/GF are
forced to deviate from the official naming rules in the same way as Intel
did.

Now, if my actual point about nodes had been discussed we would not be going in circles.