Intel kills off the 10nm process - For real??????

[R0H1T]

Wrong, I take their word when I have no reason to doubt it. In this case there is no reason to doubt it.

Corrected that for you. There is this thing called context. Intel's existing mobile SOC business was attempting to take ground an already saturated market. After market analysis they wisely decided to cut their losses and withdraw from a market they had already released products in to little fanfair. That is not the same as cancelling a mainline product advancement, which they haven't and would be fools to do.

That's is because my mind works on and requires a few things called evidence, fact and merit. Charlie's silly site, or anyone else, has failed to provide such on any level concerning the subject of Intel cancelling their 10nm process efforts. To me it looks like more of an attempt to harm investor confidence a few days before a market report is due. I find that very suspicious.

Patently false, Intel started their mobile (SoC) efforts with Atom, which went into tablets & phones. The Atom however wasn't a full SoC & Intel had to pay OEM's to buy a separate baseband radio & pair it with their Atom or get Intel's own modem in there. Intel paid billions to achieve that. Look up contra revenues, maybe that'll tell you something which most of us know. Sofia was a natural evolution & full SoC with baseband radio, which simply couldn't compete with QC or Mediatek, especially in the market it was aimed for. You're trying to rewrite history here, unfortunately that's not how it works.

 

[bug]

Patently false, Intel started their mobile (SoC) efforts with Atom, which went into tablets & phones. The Atom however wasn't a full SoC & Intel had to pay OEM's to buy a separate baseband radio & pair it with their Atom or get Intel's own modem in there. Intel paid billions to achieve that. Look up contra revenues, maybe that'll tell you something which most of us know. Sofia was a natural evolution & full SoC with baseband radio, which simply couldn't compete with QC or Mediatek, especially in the market it was aimed for. You're trying to rewrite history here, unfortunately that's not how it works.

How the hell did we get to Intel's (now defunct) mobile business, starting from allegations on semiaccurate?

We have one guy that says one thing (no sources cited) on one hand and Intel saying the guy is wrong on the other hand. End of story.

 

[lexluthermiester]

Patently false, Intel started their mobile (SoC) efforts with Atom, which went into tablets & phones.

Incorrect. Atom was started as an entry in the budding ultra mobile business back in late 2006 with "netbook" type devices and evolved from there. Don't try to rewrite history.

The Atom however wasn't a full SoC

True. And most mobile SoC's aren't either, by design.

& Intel had to pay OEM's to buy a separate baseband & pair it with their Atom or get Intel's own modem in there.

Just like anyone else.

Intel paid billions to achieve that.

Intel has invested billions into the Atom line in total, true. However they have profited from those investment over the life of the Atom line.

maybe that'll tell you something which most of us know.

No, what you're trying to do is twist facts to support your flawed agenda based narrative.

Sofia was a natural evolution & full SoC with baseband, which simply couldn't compete with QC or Mediatek, especially in the market it was aimed for.

Which is exactly why Intel canceled it. That isn't the same as cancelling a major advancement to a mainline product group.

You're trying to rewrite history here, unfortunately that's not how it works.

Take your own advice.

We have one guy that says one thing (no sources cited) on one hand and Intel saying the guy is wrong on the other hand. End of story.

Well said.

 

[R0H1T]

How the hell did we get to Intel's (now defunct) mobile business, starting from allegations on semiaccurate?

We have one guy that says one thing (no sources cited) on one hand and Intel saying the guy is wrong on the other hand. End of story.

How about you stay out of it?

Incorrect. Atom was started as an entry in the budding ultra mobile business back in late 2006 with "netbook" type devices and evolved from there. Don't try to rewrite history.

True. And most mobile SoC's aren't either, by design.

Just like anyone else.

Intel has invested billions into the Atom line in total, true. However they have profited from those investment over the life of the Atom line.

No, what you're trying to do is twist facts to support your flawed agenda based narrative.

Which is exactly why Intel canceled it. That isn't the same as cancelling a major advancement to a mainline product group.

Take your own advice.


Well said.

So, you have reading comprehension issues right ~ Intel started their mobile (SoC) efforts with Atom, which went into tablets & phones ~ Where does it say that (first) Atom was developed for mobile or tablets?

https://www.anandtech.com/show/7048...celeron-and-pentium-brands-24ghz-z3770-leaked
https://www.anandtech.com/show/4333/intels-silvermont-a-new-atom-architecture

The atom you're talking about is long dead, has been for quite some time. I've owned an Atom phone, Asus, look that up.

Wrong again, Intel made a net loss on the Atom line since 2013, you're not even on the same plane with facts on this one.

https://www.anandtech.com/show/8257/intel-q2-2014-quarterly-earnings-analysis
https://www.anandtech.com/show/8901/intel-reports-record-q4-and-full-year-revenue
https://www.anandtech.com/show/9446/intel-announces-fiscal-year-2015-quarter-two-results

In fact the losses were so bad that they moved the contra revenues into (mobile?) R&D. This was sometime in 2015/16 when they made that major change in earnings report.

Yeah I am, how about you start questioning your own misrepresentation of facts?

 

[TheoneandonlyMrK]

Wrong, I take their word when I have no reason to doubt it. In this case there is no reason to doubt it.

Corrected that for you. There is this thing called context. Intel's existing mobile SOC business was attempting to take ground an already saturated market. After market analysis they wisely decided to cut their losses and withdraw from a market they had already released products in to little fanfair. That is not the same as cancelling a mainline product advancement, which they haven't and would be fools to do.

That is because my mind works on and requires a few things called evidence, fact and merit. Charlie's silly site, or anyone else, has failed to provide such on any level concerning the subject of Intel cancelling their 10nm process efforts. To me it looks like more of an attempt to harm investor confidence a few days before a market report is due. I find that very suspicious.

Yet again asking for facts while making unsubstantiated remarks , conclusions , hypocrisy.

 

[Papahyooie]

It's way above my paygrade, but someone linked this here once.. so I'll do it too.

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

Fair enough, but Intel's 10nm process having a similar density (millions of transistors per square mm) to others' 7nm process does not make Intel's part "on par" with the 7nm parts from a gate measurement standpoint. Gate width is not the same thing as density. To be "on par" they'd have to be the same measurement (e.g, 7nm.) They may be on par in density, but not in gate measurement. Being on par in density doesn't magically make a 10nm gate be 7nm. Intel has novel designs that allow them to cram similar numbers of 10nm transistors per mm as others' processes do with 7nm, but that doesn't make Intel's part 7nm.

As such, if Intel simply called their 10nm parts 7nm, I'd contend that'd be blatant false advertising.

 

[StrayKAT]

As such, if Intel simply called their 10nm parts 7nm, I'd contend that'd be blatant false advertising.

That's true, but it also sounds like fud when tech news sites try to make it out like they're way behind in tech or something. And then dummies on Wall Street take this as gospel and create lack of confidence in the market.

I admit, Intel makes it easy for them to spread that.. but the reality sounds more complicated.

 

[Papahyooie]

That's true, but it also sounds like fud when tech news sites try to make it out like they're way behind in tech or something. And then dummies on Wall Street take this as gospel and create lack of confidence in the market.

I admit, Intel makes it easy for them to spread that.. but the reality sounds more complicated.

Oh yea, I wasn't commenting on the FUD at all... makes no difference to me. As we've pointed out here... process fab is FAR less important than people make it out to be, as far as performance is concerned. So I really don't care about the OP at all. I've just seen the "intel will just change the name to 7nm" multiple times now, and I was wondering what I was missing lol. Intel can't just simply call it 7nm if it's not 7nm, density being on par or not. (Not and still be truthful anyway.)

The OP is a moot point anyway... Intel already said it's not true, so.... case closed.

 

[TheoneandonlyMrK]

Let's just see what intel releases by next fall.

Then we'll know who's full of it.

 

[Vayra86]

If Intel says they're still working on the tech and some small guy from some little website says they're not, who's more credible? Hmm?

Yeah, I'm going to take Intel's word on their own R&D over that of some clueless nobody. All Day Long.

As always, you need to look at one question alone, and apply it to everything:

'Who benefits from this?'

Intel benefits from saying 10nm is alive and going well. Why? Because shareholders like to hear that.

Charlie does not really benefit from saying 10nm is dead. He gets a headline for it, but if its wrong, he will suffer in his credibility.

So... you make up the balance of that... At the very least, blindly believing an Intel statement is not really the most plausible choice.

 

[Papahyooie]

As always, you need to look at one question alone, and apply it to everything:
'Who benefits from this?'
Intel benefits from saying 10nm is alive and going well. Why? Because shareholders like to hear that.
Charlie does not really benefit from saying 10nm is dead. He gets a headline for it, but if its wrong, he will suffer in his credibility.
So... you make up the balance of that... At the very least, blindly believing an Intel statement is not really the most plausible choice.

While I couldn't care either way... You're saying that because Intel would benefit from saying what they said, it absolutely must be a lie.
And you're saying that Charlie doesn't benefit from saying what he said, so it must absolutely be true.

I shouldn't have to point out the logical fallacy here. Benefit or lack thereof does not equate with truth or lack thereof. Does benefit often contribute to a decision to lie? Absolutely. But lacking any further evidence, to say that Intel must inherently be lying simply because they would benefit from lying is silly.

 

[bug]

As always, you need to look at one question alone, and apply it to everything:

'Who benefits from this?'

Intel benefits from saying 10nm is alive and going well. Why? Because shareholders like to hear that.

Charlie does not really benefit from saying 10nm is dead. He gets a headline for it, but if its wrong, he will suffer in his credibility.

So... you make up the balance of that... At the very least, blindly believing an Intel statement is not really the most plausible choice.

Charlie gets clicks which is what pays his bills.
And he doesn't need credibility, he's put it in the site's name that he expects to be wrong from time to time.

As for lying to your investors, you're going to jail if you do that. Plus, what would lying achieve if Charlie was right and the process was actually killed? At some point they'd still have to admit it, but at that point they'd also have to tell their investors they burnt even more money on the process before abandoning it?

At the end of the day, there's no discussion here. Neither party has presented arguments, they've both only brought their own credibility to support the statements. All we're doing now is choosing sides in a thread that should have been long closed, but it's not because it rakes in page clicks.