(Semiwiki) Intel 14nm Delayed Again?

[Torn Mind]

I would not be surprised if there were real issues in producing those chips. Things involving highly advanced chemistry always has the potential to mess things up.

 

[Bateluer]

Given the performance of Haswell 22nm relative to AMD's Kaveri, Kabini, and Richland parts, I seriously doubt Intel is in any rush to get 14nm parts on the market.

 

[Homeles]

Given the performance of Haswell 22nm relative to AMD's Kaveri, Kabini, and Richland parts, I seriously doubt Intel is in any rush to get 14nm parts on the market.

Why do people still believe that AMD is even relevant? It's not Intel vs. AMD anymore, and it hasn't been for several years. The real topic of conversation is Intel vs. ARM, and new process nodes are extremely important there.

 

[Torn Mind]

Given the performance of Haswell 22nm relative to AMD's Kaveri, Kabini, and Richland parts, I seriously doubt Intel is in any rush to get 14nm parts on the market.

Qualcomm Snapdragon 800 series and their successors are the competition for Intel, and they need 14 nm or else they'll be left in the dust. Being the monopoly in a shrinking market that is becoming less and less profitable is not exactly the best situation to be in, and certainly not one to slow down the process node, unless the company deliberately wants to commit suicide.

 

[OCGuy]

Why do people still believe that AMD is even relevant? It's not Intel vs. AMD anymore, and it hasn't been for several years. The real topic of conversation is Intel vs. ARM, and new process nodes are extremely important there.

Who said 14nm Atom was delayed? This thread is about desktop.

 

[jdubs03]

Who said 14nm Atom was delayed? This thread is about desktop.

it's more about all 14nm releases being delayed as opposed to just desktop. but i think the themes here have already been covered.

i did read an article (rumor/prediction) from ic insights saying fab 42 will allocate capacity for apple at 10nm. i think now we're at the point where intel may have to rethink their foundry strategy, in regards to enabling competitors. adding apple will add $4 billion in revenue, not too shabby.

 

[krumme]

it's more about all 14nm releases being delayed as opposed to just desktop. but i think the themes here have already been covered.

i did read an article (rumor/prediction) from ic insights saying fab 42 will allocate capacity for apple at 10nm. i think now we're at the point where intel may have to rethink their foundry strategy, in regards to enabling competitors. adding apple will add $4 billion in revenue, not too shabby.

Its a damn difficult position. Letting Apple inside and they just steal stael steal. Inventing new technology is expensive and take years. Apple knows that. Inventions and innovation on the business skde is far mlre profitable. It must be a pain to let apple steal stuff they earned to "easy". Apple is so big and competent they can steal. Letting someone like them into your fabs is long term suiside. I would rather burn the fab and kill myself than letting leaches like apple take my stuff

But looking at intel capex and shrinking profit some radical thing must be done.

Here is my take on it if i was intel.
Kill all the not profitable in the portfolio eg the mobile drain.
Cut cost.
Kill every remotely competitor where they have stars and cashcows eg servers desktop and notebooks. It means eradiating amd now.
Kill every remotely competitor on serverside.
Make sure you are the only player.
Keep process node up at speed.
Act with paranoia. Think otellini times 5. A snake ! Lol.
Wait for the context to change then hit with your machine and cash. Hope the need for expensive high perf process will come. Look for opportunities here.

 

[ShintaiDK]

Who said 14nm Atom was delayed? This thread is about desktop.

But there is only 2 14nm chips for desktops before Skylake. And those are not delayed, because originally they was never meant to be.

 

[StrangerGuy]

Its a damn difficult position. Letting Apple inside and they just steal stael steal. Inventing new technology is expensive and take years. Apple knows that. Inventions and innovation on the business skde is far mlre profitable. It must be a pain to let apple steal stuff they earned to "easy". Apple is so big and competent they can steal. Letting someone like them into your fabs is long term suiside. I would rather burn the fab and kill myself than letting leaches like apple take my stuff

But looking at intel capex and shrinking profit some radical thing must be done.

Here is my take on it if i was intel.
Kill all the not profitable in the portfolio eg the mobile drain.
Cut cost.
Kill every remotely competitor where they have stars and cashcows eg servers desktop and notebooks. It means eradiating amd now.
Kill every remotely competitor on serverside.
Make sure you are the only player.
Keep process node up at speed.
Act with paranoia. Think otellini times 5. A snake ! Lol.
Wait for the context to change then hit with your machine and cash. Hope the need for expensive high perf process will come. Look for opportunities here.

Making Apple chips inside Intel fabs is as good as Intel admitting their own chips is no longer the king of the hill. Can't imagine anybody as arrogant as Intel would let such a thing happen.

 

[Qwertilot]

Not that extreme It'd simply involve acknowledging that another mega corp, with just a little of its own arrogance!, isn't ever going to want to surrender their tight control over their CPU design.

So long as Apple are selling this sort of stuff at the same sort of mindboggling scales they're currently doing that's probably true too.

 

[bullzz]

@StrangerGuy - arrogance was present when they were winning hands down. with profits stagnating, fabbing for apple is quick way to get around

 

[tweakboy]

1nm im wairting for that

 

[Homeles]

Arrogant isn't really a label I'd apply to Intel. That's one I'd save for Nvidia.

 

[witeken]

Here is my take on it if i was intel.
Kill all the not profitable in the portfolio eg the mobile drain.
Cut cost.
Kill every remotely competitor where they have stars and cashcows eg servers desktop and notebooks. It means eradiating amd now.
Kill every remotely competitor on serverside.
Make sure you are the only player.
Keep process node up at speed.
Act with paranoia. Think otellini times 5. A snake ! Lol.
Wait for the context to change then hit with your machine and cash. Hope the need for expensive high perf process will come. Look for opportunities here.

I wouldn't kill mobile...

Here are some optimistic ideas for what Intel could do: Intel upside potential

 

[Homeles]

Killing mobile is about the dumbest decision Intel could ever make. Forum armchair CEOs are so entertaining to observe.

 

[witeken]

I just stumbled across this interesting post on Seeking Alpha (http://seekingalpha.com/article/2014501-intel-is-this-rumor-true):

Ashraf, you are looking at Intel's comments from Investor meeting about 3 months ago.
Look at Intel's most recent earnings call transcript in January of this year. Like only 3 weeks ago.

Intel CEO comments:

"We also exited the year having made important strides on our 14-nanometer process technology. Yield improved significantly in Q4, putting us squarely on track to begin Broadwell production later this quarter. Our customers and partners also continue to evolve the computing experience. By the time we enter the back to school selling season, we'll have nearly 70 unique Tier 1 designs with outstanding battery life and performance across a range of price points and form factors."

I think you are trying a little too hard to be disappointed at Intel's advancements.
Give Intel a little more credit. And give them a little more time to crush the competition.

Seems clear to me.

Edit:
Intel's official reaction: ntel: The Truth Behind The Rumor

We continue to make progress with the industry's first 14nm manufacturing process and our second generation 3-D transistors. Broadwell, the first product on 14nm, is up-and-running as we demonstrated at the Intel Developer Forum in Q3'13. We're now planning to begin production this quarter with shipments to customers later this year.

 

[krumme]

I just stumbled across this interesting post on Seeking Alpha (http://seekingalpha.com/article/2014501-intel-is-this-rumor-true):



Seems clear to me.

Edit:
Intel's official reaction: ntel: The Truth Behind The Rumor

For some magic reason Intel is always on schedule

 

[krumme]

I wouldn't kill mobile...

Here are some optimistic ideas for what Intel could do: Intel upside potential

Yes and the ideas build on their typical cash cows earning for the rest of the business. Hardly innovating just new words.

You cant do it all. High process cost and continuing pouring money into mobile.

 

[krumme]

Killing mobile is about the dumbest decision Intel could ever make. Forum armchair CEOs are so entertaining to observe.

Beeing CEO have advantages

 

[witeken]

For some magic reason Intel is always on schedule

Not too long ago Intel actually reported that they weren't on schedule.

 

[krumme]

Not too long ago Intel actually reported that they weren't on schedule.

They rephrased it differently - but anyway they are reducing capex so i dont think its to bad for them, its a sort of funny coincidense.

 

[Homeles]

Yes and the ideas build on their typical cash cows earning for the rest of the business. Hardly innovating just new words.

You cant do it all. High process cost and continuing pouring money into mobile.

What? Yes they can. There's this thing called money, and Intel has lots of it -- plenty enough to do both. Even if they didn't have the money, they are consistent enough in their delivery of profits to allow them to borrow more.

Your precious AMD, for example, does not have money and has trouble making more of it. That's why they abandoned their foundry. That's why their S&P credit rating is crap.

What Intel can't do is continue pouring money down the drain forever. They're not doing that though, unlike what you believe. 14nm is definitely burning a hole in Intel's wallet, and they know it, and everyone knows it. However, it is an investment that will eventually pay itself off. It doesn't matter how high the up front costs are in the end -- the lower variable costs will pay for it; it's just a matter of when.

Their mobile investments are already on target to pay themselves off. I'd imagine they'd be in the black 1H 2015.

 

[VengenceIsMine]

Just like Semiaccurate?

Hardly. Charlie may play his favorites on occasion & I don't always agree with the conclusions he draws but he usually has a sound engineering basis for his rants.

 

[VengenceIsMine]

They rephrased it differently - but anyway they are reducing capex so i dont think its to bad for them, its a sort of funny coincidense.

Capex was always going to be affected by ASML bailing on 450.

 

[witeken]

Hardly. Charlie may play his favorites on occasion & I don't always agree with the conclusions he draws but he usually has a sound engineering basis for his rants.

I don't visit any of those 2 sites. It seems that I mixed up Semiaccurate and Semiwiki.